Bind ERP

binderp

Ozempic: How Useful Is AlphaFold - AlphaProteo to Easing Medication Side-Effects?

Authors:
(1) Tinghui Ouyang, National Institute of Informatics, Japan (thouyang@nii.ac.jp);
(2) Isao Echizen, National Institute of Informatics, Japan (iechizen@nii.ac.jp);
(3) Yoshiki Seo, Digital Architecture Research Center, National Institute of Advanced Industrial Science and Technology, Japan (y.seo@aist.go.jp). Table of Links Abstract and Introduction Description and Related Work of OOD Detection Methodology Experiments and Results Conclusions, Acknowledgement and References V. CONCLUSIONS In this paper, we addressed the issue of out-of-distribution (OOD) data in AI quality management by proposing a framework that combines deep learning and statistical measures for OOD detection. Initially, we leveraged the strong feature representation and dimensionality reduction capabilities of AE to extract activation traces from hidden neurons as input features for OOD analysis. Subsequently, we employed five statistical measures, namely KD, LOF, MD, kNN, and the proposed LCP, using these representative features for OOD detection. Our findings revealed that the proposed LCP, which incorporates both neighbor information and data reconstruction error, outperforms the other measures in OOD detection. Thus, it was proved to be valuable for describing and detecting OOD data. Furthermore, this research extracted reasonable corner case data with high OOD scores from the given datasets, such as MNIST, CIFAR10, and GTSRB datasets. These corner case data have high OOD scores and exhibit abnormal characteristics compared to normal data. Therefore, detecting such data using the proposed method is helpful in future AI quality assurance, particularly for quality analysis related to data security. VI. ACKNOWLEDGEMENT This research is supported by the New Energy and Industrial Technology Development Organization (NEDO) project ’JPNP20006’, and JSPS Grant-in-Aid for Early-Career Scientists (Grant Number 22K17961). REFERENCES [1] The National Institute of Advanced Industrial Science and Technology (AIST). Machine learning quality management guideline, 2023. https://www.digiarc.aist.go.jp/en/publication/aiqm/. [2] Kexin Pei, Yinzhi Cao, Junfeng Yang, and Suman Jana. Deepxplore: Automated whitebox testing of deep learning systems. In proceedings of the 26th Symposium on Operating Systems Principles, pages 1–18, 2017. [3] Tinghui Ouyang, Yoshinao Isobe, Saima Sultana, Yoshiki Seo, and Yutaka Oiwa. Autonomous driving quality assurance with data uncertainty analysis. In 2022 International Joint Conference on Neural Networks (IJCNN), pages 1–7. IEEE, 2022. [4] Guansong Pang, Jundong Li, Anton van den Hengel, Longbing Cao, and Thomas G Dietterich. Anomaly and novelty detection, explanation, and accommodation (andea). In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pages 4145–4146, 2021. [5] Jingkang Yang, Kaiyang Zhou, Yixuan Li, and Ziwei Liu. Generalized out-of-distribution detection: A survey. arXiv preprint arXiv:2110.11334, 2021. [6] Varun Chandola, Arindam Banerjee, and Vipin Kumar. Anomaly detection: A survey. ACM computing surveys (CSUR), 41(3):1–58, 2009. [7] Marco AF Pimentel, David A Clifton, Lei Clifton, and Lionel Tarassenko. A review of novelty detection. Signal processing, 99:215–249, 2014. [8] Chuanxing Geng, Sheng-jun Huang, and Songcan Chen. Recent advances in open set recognition: A survey. IEEE transactions on pattern analysis and machine intelligence, 43(10):3614–3631, 2020. [9] R´emi Domingues, Maurizio Filippone, Pietro Michiardi, and Jihane Zouaoui. A comparative evaluation of outlier detection algorithms: Experiments and analyses. Pattern recognition, 74:406–421, 2018. [10] Jing Li, Pengbo Lv, Huijun Li, and Wanghu Chen. Outlier detection based on stacked autoencoder and gaussian mixture model. In 2022 IEEE International Conference on Big Data (Big Data), pages 3763–3769. IEEE, 2022. [11] Michael E Tipping and Christopher M Bishop. Probabilistic principal component analysis. Journal of the Royal Statistical Society Series B: Statistical Methodology, 61(3):611–622, 1999. [12] John A Quinn and Masashi Sugiyama. A least-squares approach to anomaly detection in static and sequential data. Pattern Recognition Letters, 40:36–40, 2014. [13] Reuben Feinman, Ryan R Curtin, Saurabh Shintre, and Andrew B Gardner. Detecting adversarial samples from artifacts. arXiv preprint arXiv:1703.00410, 2017. [14] Kimin Lee, Kibok Lee, Honglak Lee, and Jinwoo Shin. A simple unified framework for detecting out-of-distribution samples and adversarial attacks. Advances in neural information processing systems, 31, 2018. [15] Kathrin Grosse, Praveen Manoharan, Nicolas Papernot, Michael Backes, and Patrick McDaniel. On the (statistical) detection of adversarial examples. arXiv preprint arXiv:1702.06280, 2017. [16] Ahmed Abusnaina, Yuhang Wu, Sunpreet Arora, Yizhen Wang, Fei Wang, Hao Yang, and David Mohaisen. Adversarial example detection using latent neighborhood graph. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 7687–7696, 2021. [17] Bruno Andriamanalimanana, Ali Tekeoglu, Korkut Bekiroglu, Saumendra Sengupta, Chen-Fu Chiang, Michael Reale, and Jorge Novillo. Symmetric kullback-leibler divergence of softmaxed distributions for anomaly scores. In 2019 IEEE Conference on Communications and Network Security (CNS), pages 1–6. IEEE, 2019. [18] Jianbo Yu, Xiaoyun Zheng, and Shijin Wang. A deep autoencoder feature learning method for process pattern recognition. Journal of Process Control, 79:1–15, 2019. [19] Lukas Ruff, Robert Vandermeulen, Nico Goernitz, Lucas Deecke, Shoaib Ahmed Siddiqui, Alexander Binder, Emmanuel M¨uller, and Marius Kloft. Deep one-class classification. In International conference on machine learning, pages 4393–4402. PMLR, 2018. [20] Jinhan Kim, Robert Feldt, and Shin Yoo. Guiding deep learning system testing using surprise adequacy. In 2019 IEEE/ACM 41st International Conference on Software Engineering (ICSE), pages 1039–1049. IEEE, 2019. [21] Zenan Li, Xiaoxing Ma, Chang Xu, and Chun Cao. Structural coverage criteria for neural networks could be misleading. In 2019 IEEE/ACM 41st International Conference on Software Engineering: New Ideas and Emerging Results (ICSE-NIER), pages 89–92. IEEE, 2019. [22] Shiqing Ma, Yingqi Liu, Guanhong Tao, Wen-Chuan Lee, and Xiangyu Zhang. Nic: Detecting adversarial samples with neural network invariant checking. In 26th Annual Network And Distributed System Security Symposium (NDSS 2019). Internet Soc, 2019. [23] Tarem Ahmed. Online anomaly detection using kde. In GLOBECOM 2009-2009 IEEE Global Telecommunications Conference, pages 1–8. IEEE, 2009. [24] Siying Xu, Huiyi Liu, Liting Duan, andWenjingWu. An improved lof outlier detection algorithm. In 2021 IEEE International Conference on Artificial Intelligence and Computer Applications (ICAICA), pages 113–117. IEEE, 2021. [25] Laurens Van der Maaten and Geoffrey Hinton. Visualizing data using t-sne. Journal of machine learning research, 9(11), 2008. [26] Yann LeCun. The mnist database of handwritten digits. http://yann. lecun. com/exdb/mnist/, 1998. [27] Alex Krizhevsky, Geoffrey Hinton, et al. Learning multiple layers of features from tiny images. 2009. [28] Tom Fawcett. An introduction to roc analysis. Pattern recognition letters, 27(8):861–874, 2006. [29] Tinghui Ouyang, Vicent Sanz Marco, Yoshinao Isobe, Hideki Asoh, Yutaka Oiwa, and Yoshiki Seo. Corner case data description and detection. In 2021 IEEE/ACM 1st Workshop on AI Engineering-Software Engineering for AI (WAIN), pages 19–26. IEEE, 2021. [30] Stallkamp, Johannes, Marc Schlipsing, Jan Salmen, and Christian Igel. The German traffic sign recognition benchmark: a multi-class classification competition. In The 2011 international joint conference on neural networks, pages 1453-1460. IEEE, 2011. This paper is available on arxiv under CC 4.0 license. Authors: (1) Tinghui Ouyang, National Institute of Informatics, Japan (thouyang@nii.ac.jp); (2) Isao Echizen, National Institute of Informatics, Japan (iechizen@nii.ac.jp); (3) Yoshiki Seo, Digital Architecture Research Center, National Institute of Advanced Industrial Science and Technology, Japan (y.seo@aist.go.jp). Authors: Authors: (1) Tinghui Ouyang, National Institute of Informatics, Japan (thouyang@nii.ac.jp); (2) Isao Echizen, National Institute of Informatics, Japan (iechizen@nii.ac.jp); (3) Yoshiki Seo, Digital Architecture Research Center, National Institute of Advanced Industrial Science and Technology, Japan (y.seo@aist.go.jp). Table of Links Abstract and Introduction Abstract and Introduction Description and Related Work of OOD Detection Description and Related Work of OOD Detection Methodology Methodology Experiments and Results Experiments and Results Conclusions, Acknowledgement and References Conclusions, Acknowledgement and References V. CONCLUSIONS In this paper, we addressed the issue of out-of-distribution (OOD) data in AI quality management by proposing a framework that combines deep learning and statistical measures for OOD detection. Initially, we leveraged the strong feature representation and dimensionality reduction capabilities of AE to extract activation traces from hidden neurons as input features for OOD analysis. Subsequently, we employed five statistical measures, namely KD, LOF, MD, kNN, and the proposed LCP, using these representative features for OOD detection. Our findings revealed that the proposed LCP, which incorporates both neighbor information and data reconstruction error, outperforms the other measures in OOD detection. Thus, it was proved to be valuable for describing and detecting OOD data. Furthermore, this research extracted reasonable corner case data with high OOD scores from the given datasets, such as MNIST, CIFAR10, and GTSRB datasets. These corner case data have high OOD scores and exhibit abnormal characteristics compared to normal data. Therefore, detecting such data using the proposed method is helpful in future AI quality assurance, particularly for quality analysis related to data security. VI. ACKNOWLEDGEMENT This research is supported by the New Energy and Industrial Technology Development Organization (NEDO) project ’JPNP20006’, and JSPS Grant-in-Aid for Early-Career Scientists (Grant Number 22K17961). REFERENCES [1] The National Institute of Advanced Industrial Science and Technology (AIST). Machine learning quality management guideline, 2023. https://www.digiarc.aist.go.jp/en/publication/aiqm/. [2] Kexin Pei, Yinzhi Cao, Junfeng Yang, and Suman Jana. Deepxplore: Automated whitebox testing of deep learning systems. In proceedings of the 26th Symposium on Operating Systems Principles, pages 1–18, 2017. [3] Tinghui Ouyang, Yoshinao Isobe, Saima Sultana, Yoshiki Seo, and Yutaka Oiwa. Autonomous driving quality assurance with data uncertainty analysis. In 2022 International Joint Conference on Neural Networks (IJCNN), pages 1–7. IEEE, 2022. [4] Guansong Pang, Jundong Li, Anton van den Hengel, Longbing Cao, and Thomas G Dietterich. Anomaly and novelty detection, explanation, and accommodation (andea). In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pages 4145–4146, 2021. [5] Jingkang Yang, Kaiyang Zhou, Yixuan Li, and Ziwei Liu. Generalized out-of-distribution detection: A survey. arXiv preprint arXiv:2110.11334, 2021. [6] Varun Chandola, Arindam Banerjee, and Vipin Kumar. Anomaly detection: A survey. ACM computing surveys (CSUR), 41(3):1–58, 2009. [7] Marco AF Pimentel, David A Clifton, Lei Clifton, and Lionel Tarassenko. A review of novelty detection. Signal processing, 99:215–249, 2014. [8] Chuanxing Geng, Sheng-jun Huang, and Songcan Chen. Recent advances in open set recognition: A survey. IEEE transactions on pattern analysis and machine intelligence, 43(10):3614–3631, 2020. [9] R´emi Domingues, Maurizio Filippone, Pietro Michiardi, and Jihane Zouaoui. A comparative evaluation of outlier detection algorithms: Experiments and analyses. Pattern recognition, 74:406–421, 2018. [10] Jing Li, Pengbo Lv, Huijun Li, and Wanghu Chen. Outlier detection based on stacked autoencoder and gaussian mixture model. In 2022 IEEE International Conference on Big Data (Big Data), pages 3763–3769. IEEE, 2022. [11] Michael E Tipping and Christopher M Bishop. Probabilistic principal component analysis. Journal of the Royal Statistical Society Series B: Statistical Methodology, 61(3):611–622, 1999. [12] John A Quinn and Masashi Sugiyama. A least-squares approach to anomaly detection in static and sequential data. Pattern Recognition Letters, 40:36–40, 2014. [13] Reuben Feinman, Ryan R Curtin, Saurabh Shintre, and Andrew B Gardner. Detecting adversarial samples from artifacts. arXiv preprint arXiv:1703.00410, 2017. [14] Kimin Lee, Kibok Lee, Honglak Lee, and Jinwoo Shin. A simple unified framework for detecting out-of-distribution samples and adversarial attacks. Advances in neural information processing systems, 31, 2018. [15] Kathrin Grosse, Praveen Manoharan, Nicolas Papernot, Michael Backes, and Patrick McDaniel. On the (statistical) detection of adversarial examples. arXiv preprint arXiv:1702.06280, 2017. [16] Ahmed Abusnaina, Yuhang Wu, Sunpreet Arora, Yizhen Wang, Fei Wang, Hao Yang, and David Mohaisen. Adversarial example detection using latent neighborhood graph. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 7687–7696, 2021. [17] Bruno Andriamanalimanana, Ali Tekeoglu, Korkut Bekiroglu, Saumendra Sengupta, Chen-Fu Chiang, Michael Reale, and Jorge Novillo. Symmetric kullback-leibler divergence of softmaxed distributions for anomaly scores. In 2019 IEEE Conference on Communications and Network Security (CNS), pages 1–6. IEEE, 2019. [18] Jianbo Yu, Xiaoyun Zheng, and Shijin Wang. A deep autoencoder feature learning method for process pattern recognition. Journal of Process Control, 79:1–15, 2019. [19] Lukas Ruff, Robert Vandermeulen, Nico Goernitz, Lucas Deecke, Shoaib Ahmed Siddiqui, Alexander Binder, Emmanuel M¨uller, and Marius Kloft. Deep one-class classification. In International conference on machine learning, pages 4393–4402. PMLR, 2018. [20] Jinhan Kim, Robert Feldt, and Shin Yoo. Guiding deep learning system testing using surprise adequacy. In 2019 IEEE/ACM 41st International Conference on Software Engineering (ICSE), pages 1039–1049. IEEE, 2019. [21] Zenan Li, Xiaoxing Ma, Chang Xu, and Chun Cao. Structural coverage criteria for neural networks could be misleading. In 2019 IEEE/ACM 41st International Conference on Software Engineering: New Ideas and Emerging Results (ICSE-NIER), pages 89–92. IEEE, 2019. [22] Shiqing Ma, Yingqi Liu, Guanhong Tao, Wen-Chuan Lee, and Xiangyu Zhang. Nic: Detecting adversarial samples with neural network invariant checking. In 26th Annual Network And Distributed System Security Symposium (NDSS 2019). Internet Soc, 2019. [23] Tarem Ahmed. Online anomaly detection using kde. In GLOBECOM 2009-2009 IEEE Global Telecommunications Conference, pages 1–8. IEEE, 2009. [24] Siying Xu, Huiyi Liu, Liting Duan, andWenjingWu. An improved lof outlier detection algorithm. In 2021 IEEE International Conference on Artificial Intelligence and Computer Applications (ICAICA), pages 113–117. IEEE, 2021. [25] Laurens Van der Maaten and Geoffrey Hinton. Visualizing data using t-sne. Journal of machine learning research, 9(11), 2008. [26] Yann LeCun. The mnist database of handwritten digits. http://yann. lecun. com/exdb/mnist/, 1998. [27] Alex Krizhevsky, Geoffrey Hinton, et al. Learning multiple layers of features from tiny images. 2009. [28] Tom Fawcett. An introduction to roc analysis. Pattern recognition letters, 27(8):861–874, 2006. [29] Tinghui Ouyang, Vicent Sanz Marco, Yoshinao Isobe, Hideki Asoh, Yutaka Oiwa, and Yoshiki Seo. Corner case data description and detection. In 2021 IEEE/ACM 1st Workshop on AI Engineering-Software Engineering for AI (WAIN), pages 19–26. IEEE, 2021. [30] Stallkamp, Johannes, Marc Schlipsing, Jan Salmen, and Christian Igel. The German traffic sign recognition benchmark: a multi-class classification competition. In The 2011 international joint conference on neural networks, pages 1453-1460. IEEE, 2011. This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv

Advancing OOD Detection: A Deep Learning Approach

Table of Links Abstract and I. Introduction II. Expectations of Blockchain Accessibility III. Approach I: Maintain a Ledger Locally - Run a Full Node IV. Approach II: Query a Third-Party Ledger-Node-As-A-Service (NAAS) V. Approach III: Light Node - External Query & Local Verification VI. Concluding Remark and References VI. CONCLUDING REMARKS Ultimately, any blockchain accessibility solution has to decide on a split of responsibilities: either the user does most of the computation necessary or some third party does. A spectrum of solutions may be developed based on different types and needs of users, from a full node to a full third-party service. However, we consider that any solution must still adhere to reasonable security expectations: integrity, availability, and privacy. Most blockchain research is based on the assumption that users run their local nodes. This means the fact that running a full node is still the best way to provide integrity, availability, and privacy. Indeed we surveyed third-party service providers in the field and noted the lack of a data integrity guarantee when the provider is not trusted. We also recalled events that render the providers unable to provide service. These facts mean that third-party service providers can and do fail, and thus relying on them can become risky. At the same time, considering the high cost of running a full node locally, we believe that users should have more choice that does not impact the security guarantees. We observe the emergence of so-called ultralight nodes, which provides a way for third-party services to give a verifiable statement that requires little hardware power from the client. Many open questions remain regarding the accessibility of blockchains, even regarding full-node solutions and third-party providers. Any number of light node protocols can also be built to serve different users’ needs. We reiterate that ensuring the accessibility of blockchains is paramount for the technology. If we believe that blockchains are a common good, then making them accessible to all in a secure and available manner is critical for the future development of Web3 technology. Any information is as useful as how it can be read, and such reasoning also applies to blockchains. REFERENCES [1] AL-BASSAM, M., SONNINO, A., AND BUTERIN, V. Fraud and data availability proofs: Maximising light client security and scaling blockchains with dishonest majorities. arXiv preprint arXiv:1809.09044 (2018). [2] ALCHEMY. Alchemy. https://alchemy.com/, 2022. [3] ALL THAT NODE. All That Node. https://allthatnode.com/, 2022. [4] ANKR. Ankr. https://www.ankr.com/, 2022. [5] BADDOUR, R. Setting up a Litecoin super node: A complete step-bystep recipe that anyone can follow. https://medium.com/@baddour/sett ing-up-a-litecoin-full-node-a-complete-step-by-step-recipe-that-anyon e-can-follow-fb18b82d5cfe, 2022. [6] BINANCE. Binance. https://www.binance.us/, 2022. [7] BITCOIN PROJECT. Minimum requirements. https://bitcoin.org/en/ful l-node#minimum-requirements, 2022. [8] BITCOIN PROJECT. RPC API reference. https://developer.bitcoin.org/ reference/rpc/#, 2022. [9] BLOCKDAEMON. BlockDeamon. https://blockdaemon.com/, 2022. [10] BUNZ, B., KIFFER, L., LUU, L., AND ZAMANI, M. Flyclient: Superlight clients for cryptocurrencies. In 2020 IEEE Symposium on Security and Privacy (SP) (2020), IEEE, pp. 928–946. [11] CHAINSTACK. Chainstack. https://chainstack.com/, 2022. [12] CHATZIGIANNIS, P., BALDIMTSI, F., AND CHALKIAS, K. Sok: Blockchain light clients. In International Conference on Financial Cryptography and Data Security (2022), Springer, pp. 615–641. [13] CLEPER, Y., AND BINDER, G. Lava: A decentralized RPC network for blockchains - litepaper v0.1. Tech. rep., Lava Protocol Inc., 2022. [14] COINCASHEW. How to run your own Monero node. https://www.co incashew.com/coins/overview-xmr/guide-or-how-to-run-a-full-node, 2022. [15] CONSENSYS. Privacy policy. http://web.archive.org/web/2022112109 5716/https://consensys.net/privacy-policy/, 2022. [16] CRYPTO.COM. Crypto.com. https://crypto.com/us, 2022. [17] DAIAN, P., GOLDFEDER, S., KELL, T., LI, Y., ZHAO, X., BENTOV, I., BREIDENBACH, L., AND JUELS, A. Flash boys 2.0: Frontrunning in decentralized exchanges, miner extractable value, and consensus instability. In 2020 IEEE Symposium on Security and Privacy (SP) (2020), IEEE, pp. 910–927. [18] DASH CORE GROUP. Masternode requirements. https://docs.dash.org/ en/stable/masternodes/understanding.html#masternode-requirements, 2022. [19] DATAHUB. DataHub. https://datahub.figment.io/, 2022. [20] DENNIS, R., AND DISSO, J. P. An analysis into the scalability of bitcoin and ethereum. In Third International Congress on Information and Communication Technology (2019), Springer, pp. 619–627. [21] ELECTRIC COIN COMPANY. Troubleshooting guide. https://zcash.read thedocs.io/en/latest/rtd pages/troubleshooting guide.html, 2022. [22] ETHEREUM FOUNDATION. Ethereum portal network. https://www.ethp ortal.net/, 2022. [23] ETHEREUM FOUNDATION. Light Ethereum subprotocol (LES). https: //github.com/ethereum/devp2p/blob/master/caps/les.md, 2022. [24] ETHEREUM FOUNDATION. Light node. https://ethereum.org/en/develo pers/docs/nodes-and-clients/#light-node, 2022. [25] ETHEREUM FOUNDATION. Node as a service. https://ethereum.org/en/ developers/docs/nodes-and-clients/nodes-as-a-service/, 2022. [26] ETHEREUM FOUNDATION. Requirements. https://ethereum.org/en/deve lopers/docs/nodes-and-clients/run-a-node/#requirements, 2022. [27] ETHEREUM FOUNDATION. Who should run a node? https://ethereum .org/en/run-a-node/, 2022. [28] FANTI, G., KOGAN, L., AND VISWANATH, P. Economics of proof-ofstake payment systems, 2019. [29] GETBLOCK. GetBlock. https://getblock.io/, 2022. [30] GOES, C. The interblockchain communication protocol: An overview. arXiv preprint arXiv:2006.15918 (2020). [31] GRIFFITH, E. Why the crypto collapse matters. The New York Times (Nov 2022). [32] INFSTONES. InfStones. https://infstones.com/, 2022. [33] INFURA. Infura. https://infura.io/, 2022. [34] KALEIDO. Kaleido. https://kaleido.io/, 2022. [35] LI, K., CHEN, J., LIU, X., TANG, Y. R., WANG, X., AND LUO, X. As strong as its weakest link: How to break blockchain dapps at rpc service. In NDSS (2021). [36] MALWA, S. Two Polygon, Fantom front ends hit by DNS attack. https: //www.coindesk.com/tech/2022/07/01/two-polygon-fantom-front-end s-hit-by-dns-attack/, 2022. [37] MEE, D. Bitcoin full node on RBP3 (revised). https://medium.com/@ meeDamian/bitcoin-full-node-on-rbp3-revised-88bb7c8ef1d1, 2017. [38] MEE, D. Litecoin full node on RBP3. https://medium.com/@meeDam ian/litecoin-full-node-on-rbp3-69f851f74f1b, 2017. [39] METAMASK SUPPORT. Why Infura cannot serve certain areas. https: //metamask.zendesk.com/hc/en-us/articles/360059386712-Why-MetaM ask-and-Infura-cannot-serve-certain-areas, 2022. [40] MORALIS. Moralis. https://moralis.io/, 2022. [41] NAKAMOTO, S. Bitcoin: A peer-to-peer electronic cash system. Decentralized Business Review (2008), 21260. [42] NELSON, B. J. Remote procedure call. PhD thesis, Carnegie Mellon University, 1981. [43] NOWNODES. NOWNodes. https://nownodes.io/, 2022. [44] PAAVOLAINEN, S., AND CARR, C. Security properties of light clients on the ethereum blockchain. IEEE Access 8 (2020), 124339–124358. [45] POCKET NETWORK. Pocket Network. https://www.pokt.network/, 2022. [46] QUICKNODE. QuickNode. https://www.quicknode.com/, 2022. [47] RIVET. Rivet. https://rivet.cloud/, 2022. [48] SENSEINODE. SenseiNode. https://senseinode.com/, 2022. [49] SETTLEMINT. SettleMint. https://console.settlemint.com/, 2022. [50] SOLANA FOUNDATION. Validator requirements. https://docs.solana.co m/running-validator/validator-reqs, 2022. [51] SONG, Y.-D., AND ASTE, T. The cost of bitcoin mining has never really increased. Frontiers in Blockchain (2020), 44. [52] WATCHDATA. WatchData. https://watchdata.io/, 2022. [53] XRP LEDGER. System requirements. https://xrpl.org/system-requireme nts.html, 2022. [54] YAISH, A., AND ZOHAR, A. Correct cryptocurrency asic pricing: Are miners overpaying. arXiv preprint arXiv:2002.11064 (2020). [55] ZMOK. ZMOK. https://zmok.io/, 2022. Authors:
(1) Zhongtang Luo, Purdue University (luo401@purdue.edu);
(2) Rohan Murukutla, Supra (r.murukutla@supraoracles.com);
(3) Aniket Kate, Purdue University / Supra (aniket@purdue.edu). This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license. Table of Links Abstract and I. Introduction Abstract and I. Introduction II. Expectations of Blockchain Accessibility II. Expectations of Blockchain Accessibility III. Approach I: Maintain a Ledger Locally - Run a Full Node III. Approach I: Maintain a Ledger Locally - Run a Full Node IV. Approach II: Query a Third-Party Ledger-Node-As-A-Service (NAAS) IV. Approach II: Query a Third-Party Ledger-Node-As-A-Service (NAAS) V. Approach III: Light Node - External Query & Local Verification V. Approach III: Light Node - External Query & Local Verification VI. Concluding Remark and References VI. Concluding Remark and References VI. CONCLUDING REMARKS Ultimately, any blockchain accessibility solution has to decide on a split of responsibilities: either the user does most of the computation necessary or some third party does. A spectrum of solutions may be developed based on different types and needs of users, from a full node to a full third-party service. However, we consider that any solution must still adhere to reasonable security expectations: integrity, availability, and privacy. Most blockchain research is based on the assumption that users run their local nodes. This means the fact that running a full node is still the best way to provide integrity, availability, and privacy. Indeed we surveyed third-party service providers in the field and noted the lack of a data integrity guarantee when the provider is not trusted. We also recalled events that render the providers unable to provide service. These facts mean that third-party service providers can and do fail, and thus relying on them can become risky. At the same time, considering the high cost of running a full node locally, we believe that users should have more choice that does not impact the security guarantees. We observe the emergence of so-called ultralight nodes, which provides a way for third-party services to give a verifiable statement that requires little hardware power from the client. Many open questions remain regarding the accessibility of blockchains, even regarding full-node solutions and third-party providers. Any number of light node protocols can also be built to serve different users’ needs. We reiterate that ensuring the accessibility of blockchains is paramount for the technology. If we believe that blockchains are a common good, then making them accessible to all in a secure and available manner is critical for the future development of Web3 technology. Any information is as useful as how it can be read, and such reasoning also applies to blockchains. REFERENCES [1] AL-BASSAM, M., SONNINO, A., AND BUTERIN, V. Fraud and data availability proofs: Maximising light client security and scaling blockchains with dishonest majorities. arXiv preprint arXiv:1809.09044 (2018). [2] ALCHEMY. Alchemy. https://alchemy.com/, 2022. [3] ALL THAT NODE. All That Node. https://allthatnode.com/, 2022. [4] ANKR. Ankr. https://www.ankr.com/, 2022. [5] BADDOUR, R. Setting up a Litecoin super node: A complete step-bystep recipe that anyone can follow. https://medium.com/@baddour/sett ing-up-a-litecoin-full-node-a-complete-step-by-step-recipe-that-anyon e-can-follow-fb18b82d5cfe, 2022. [6] BINANCE. Binance. https://www.binance.us/, 2022. [7] BITCOIN PROJECT. Minimum requirements. https://bitcoin.org/en/ful l-node#minimum-requirements, 2022. [8] BITCOIN PROJECT. RPC API reference. https://developer.bitcoin.org/ reference/rpc/#, 2022. [9] BLOCKDAEMON. BlockDeamon. https://blockdaemon.com/, 2022. [10] BUNZ, B., KIFFER, L., LUU, L., AND ZAMANI, M. Flyclient: Superlight clients for cryptocurrencies. In 2020 IEEE Symposium on Security and Privacy (SP) (2020), IEEE, pp. 928–946. [11] CHAINSTACK. Chainstack. https://chainstack.com/, 2022. [12] CHATZIGIANNIS, P., BALDIMTSI, F., AND CHALKIAS, K. Sok: Blockchain light clients. In International Conference on Financial Cryptography and Data Security (2022), Springer, pp. 615–641. [13] CLEPER, Y., AND BINDER, G. Lava: A decentralized RPC network for blockchains - litepaper v0.1. Tech. rep., Lava Protocol Inc., 2022. [14] COINCASHEW. How to run your own Monero node. https://www.co incashew.com/coins/overview-xmr/guide-or-how-to-run-a-full-node, 2022. [15] CONSENSYS. Privacy policy. http://web.archive.org/web/2022112109 5716/https://consensys.net/privacy-policy/, 2022. [16] CRYPTO.COM. Crypto.com. https://crypto.com/us, 2022. [17] DAIAN, P., GOLDFEDER, S., KELL, T., LI, Y., ZHAO, X., BENTOV, I., BREIDENBACH, L., AND JUELS, A. Flash boys 2.0: Frontrunning in decentralized exchanges, miner extractable value, and consensus instability. In 2020 IEEE Symposium on Security and Privacy (SP) (2020), IEEE, pp. 910–927. [18] DASH CORE GROUP. Masternode requirements. https://docs.dash.org/ en/stable/masternodes/understanding.html#masternode-requirements, 2022. [19] DATAHUB. DataHub. https://datahub.figment.io/, 2022. [20] DENNIS, R., AND DISSO, J. P. An analysis into the scalability of bitcoin and ethereum. In Third International Congress on Information and Communication Technology (2019), Springer, pp. 619–627. [21] ELECTRIC COIN COMPANY. Troubleshooting guide. https://zcash.read thedocs.io/en/latest/rtd pages/troubleshooting guide.html, 2022. [22] ETHEREUM FOUNDATION. Ethereum portal network. https://www.ethp ortal.net/, 2022. [23] ETHEREUM FOUNDATION. Light Ethereum subprotocol (LES). https: //github.com/ethereum/devp2p/blob/master/caps/les.md, 2022. [24] ETHEREUM FOUNDATION. Light node. https://ethereum.org/en/develo pers/docs/nodes-and-clients/#light-node, 2022. [25] ETHEREUM FOUNDATION. Node as a service. https://ethereum.org/en/ developers/docs/nodes-and-clients/nodes-as-a-service/, 2022. [26] ETHEREUM FOUNDATION. Requirements. https://ethereum.org/en/deve lopers/docs/nodes-and-clients/run-a-node/#requirements, 2022. [27] ETHEREUM FOUNDATION. Who should run a node? https://ethereum .org/en/run-a-node/, 2022. [28] FANTI, G., KOGAN, L., AND VISWANATH, P. Economics of proof-ofstake payment systems, 2019. [29] GETBLOCK. GetBlock. https://getblock.io/, 2022. [30] GOES, C. The interblockchain communication protocol: An overview. arXiv preprint arXiv:2006.15918 (2020). [31] GRIFFITH, E. Why the crypto collapse matters. The New York Times (Nov 2022). [32] INFSTONES. InfStones. https://infstones.com/, 2022. [33] INFURA. Infura. https://infura.io/, 2022. [34] KALEIDO. Kaleido. https://kaleido.io/, 2022. [35] LI, K., CHEN, J., LIU, X., TANG, Y. R., WANG, X., AND LUO, X. As strong as its weakest link: How to break blockchain dapps at rpc service. In NDSS (2021). [36] MALWA, S. Two Polygon, Fantom front ends hit by DNS attack. https: //www.coindesk.com/tech/2022/07/01/two-polygon-fantom-front-end s-hit-by-dns-attack/, 2022. [37] MEE, D. Bitcoin full node on RBP3 (revised). https://medium.com/@ meeDamian/bitcoin-full-node-on-rbp3-revised-88bb7c8ef1d1, 2017. [38] MEE, D. Litecoin full node on RBP3. https://medium.com/@meeDam ian/litecoin-full-node-on-rbp3-69f851f74f1b, 2017. [39] METAMASK SUPPORT. Why Infura cannot serve certain areas. https: //metamask.zendesk.com/hc/en-us/articles/360059386712-Why-MetaM ask-and-Infura-cannot-serve-certain-areas, 2022. [40] MORALIS. Moralis. https://moralis.io/, 2022. [41] NAKAMOTO, S. Bitcoin: A peer-to-peer electronic cash system. Decentralized Business Review (2008), 21260. [42] NELSON, B. J. Remote procedure call. PhD thesis, Carnegie Mellon University, 1981. [43] NOWNODES. NOWNodes. https://nownodes.io/, 2022. [44] PAAVOLAINEN, S., AND CARR, C. Security properties of light clients on the ethereum blockchain. IEEE Access 8 (2020), 124339–124358. [45] POCKET NETWORK. Pocket Network. https://www.pokt.network/, 2022. [46] QUICKNODE. QuickNode. https://www.quicknode.com/, 2022. [47] RIVET. Rivet. https://rivet.cloud/, 2022. [48] SENSEINODE. SenseiNode. https://senseinode.com/, 2022. [49] SETTLEMINT. SettleMint. https://console.settlemint.com/, 2022. [50] SOLANA FOUNDATION. Validator requirements. https://docs.solana.co m/running-validator/validator-reqs, 2022. [51] SONG, Y.-D., AND ASTE, T. The cost of bitcoin mining has never really increased. Frontiers in Blockchain (2020), 44. [52] WATCHDATA. WatchData. https://watchdata.io/, 2022. [53] XRP LEDGER. System requirements. https://xrpl.org/system-requireme nts.html, 2022. [54] YAISH, A., AND ZOHAR, A. Correct cryptocurrency asic pricing: Are miners overpaying. arXiv preprint arXiv:2002.11064 (2020). [55] ZMOK. ZMOK. https://zmok.io/, 2022. Authors: (1) Zhongtang Luo, Purdue University (luo401@purdue.edu); (2) Rohan Murukutla, Supra (r.murukutla@supraoracles.com); (3) Aniket Kate, Purdue University / Supra (aniket@purdue.edu). Authors: Authors: (1) Zhongtang Luo, Purdue University (luo401@purdue.edu); (2) Rohan Murukutla, Supra (r.murukutla@supraoracles.com); (3) Aniket Kate, Purdue University / Supra (aniket@purdue.edu). This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license. This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license. available on arxiv

Balancing Blockchain Access: Local Nodes vs. Third-Party Services

Orca 2: Enhancing Reasoning in Smaller Language Models - Conclusions and References

This paper is available on arxiv under CC 4.0 license. Authors: (1) Md Masud Rana, Department of Mathematics, University of Kentucky; (2) Duc Duy Nguyen, Department of Mathematics, University of Kentucky & ducnguyen@uky.edu. Table of Links Abstract & Introduction Datasets and Results Methods Conclusion, Data and Software Availability, Competing interests, Acknowledgments & References 2 Datasets and Results In this section, we perform validation and evaluation of our proposed models on several benchmark datasets. We develop two types of GGL-PPI models: GGL-PPI1 and GGL-PPI2. The first model, GGL-PPI1, is built solely on geometric graph features discussed in Section 3. On the other hand, GGL-PPI2 incorporates both geometric graph features and auxiliary features, as detailed by Wang et al. [41]. The electrostatic potential calculations for the auxiliary components are conducted using the MIBPB software [42]. 2.1 Validation To validate our models, we primarily consider the AB-Bind dataset [25], SKEMPI 1.0 dataset [23], and SKEMPI 2.0 dataset [24]. We employ a rigorous evaluation methodology by conducting a 10-times 10-fold cross-validation (CV) on each datasets. The mean Pearson correlation coefficient (R ) and root-mean-square error (RMSE) serve as our evaluation metrics. p In comparing the CV performance of our proposed models with other existing methods, we specifically assess TopNetTree [41], Hom-ML-V2 [43], and Hom-ML-V1 [43]. Both TopNetTree and Hom-ML-V2 incorporate auxiliary features in conjunction with their topology-based and Hom-complex-based features, respectively. On the other hand, Hom-ML-V1 solely relies on Hom-complex-based features without utilizing any auxiliary features. Validation on AB-Bind S645 Data Set The AB-Bind dataset contains 1,101 mutational data points for 32 antibody-antigen complexes, providing experimentally determined binding affinity changes upon mutations. Pires et al. curated a subset known as AB-Bind S645 [44], consisting of 645 single-point mutations observed in 29 antibody-antigen complexes. The dataset comprises a mix of stabilizing (20%) and destabilizing (80%) mutations. Additionally, the dataset includes 27 non-binders that do not show any binding within the assay’s sensitivity range. For these non-binders, the binding free energy changes have been uniformly set to a value of 8 kcal/mol. It is crucial to consider these non-binders as outliers during model development and evaluation to ensure model accuracy and robustness. Our GGL-PPI2 achieved an Rp of 0.58 on the AB-Bind S645 dataset, as shown in Figure 2a. The comparison results in Table 1 indicate that our model tied for second place with Hom-ML-V2 43, while TopNetTree [41] claimed the top position. However, when we exclude the 27 nonbinders from the dataset, our model outperforms all other existing models. Specifically, the Rp value increases to 0.74 from 0.58 after removing the nonbinders (Figure 2b). Furthermore, GGL-PI1, our purely geometric graph-based features model, demonstrated competitive performance with an Rp of 0.57 on the AB-Bind S645 dataset. Intriguingly, when excluding the nonbinders, GGL-PPI1 surpassed all other models with an improved Rp of 0.73. These performances reveal that our multiscale weighted colored geometric graphs can effectively characterize the wide range of interactions in biomolecular complexes. Validation on SKEMPI 1.0 S1131 Data Set The SKEMPI 1.0 dataset consists of a collection of 3,047 mutations of 158 complexes obtained from literature sources, where the complexes have experimentally determined structures [23]. The dataset includes both single-point mutations and multi-point mutations. Specifically, there are 2,317 entries in the dataset that represent single-point mutations, which are collectively known as the SKEMPI S2317 set. Additionally, a subset of 1,131 non-redundant interface single-point mutations has been selected from the SKEMPI S2317 set and labeled as the SKEMPI S1131 set [45]. This subset focuses on studying the impact of single-point mutations on protein-protein interactions. Table 1: Performance comparison of different methods in terms of Pearson correlation coefficients (Rp) for the AB-Bind (S645) dataset. Figure 2c shows that our model GGL-PPI2 achieves an Rp of 0.873 and an RMSE of 1.21 kcal/mol in 10-fold CV on the S1131 dataset. Table 2 presents the performance comparison of various methods on the S1131 dataset, including our proposed models, GGL-PPI1 and GGL-PPI2. Among them, our model, GGL-PPI2, achieved the highest performance, underscoring its superiority in predicting binding affinity changes due to mutation. Notably, even without auxiliary features, our GGL-PPI1 outperformed both TopNetTree and Hom-ML-V2 methods that do leverage auxiliary features. This again highlights the efficacy of our geometric graph-based molecular representation. Table 2: Performance comparison of different methods in terms of Pearson correlation coefficients (Rp) for the single-point mutations in the SKEMPI 1.0 (S1131) dataset. Validation on SKEMPI 2.0 S4169 and S8338 Data Sets The SKEMPI 2.0 dataset is an updated and expanded version of the original SKEMPI dataset, incorporating new mutations collected from various sources [24]. Released in 2018, it significantly increased in size, now containing a total of 7,085 entries, including both single-point and multi-point mutations. The data was obtained by merging several databases, including SKEMPI 1.0 [23], AB-Bind [25], PROXiMATE27, and dbMPIKT46. Additionally, new data from the literature were manually curated and added to the dataset. The mutations cover a wide range of protein complexes, such as protease-inhibitor, antibody-antigen, and TRCpMHC complexes. Among the mutations, approximately 3,000 are single-point alanine mutations, 2,000 are single-point non-alanine mutations, and another 2,000 involve multiple mutations. Notably, the authors of the mCSM-PPI2 [8] method filtered the single-point mutations, yielding S4169 set, comprising 4,169 variants in 139 different complexes/ The S8338 set, derived from S4169, represents hypothetical reverse mutation energy changes with negative values. This comprehensive dataset serves as a valuable resource for studying protein interactions and their thermodynamic properties. Perforamnce-wise, Our GGL-PPI2 model posts an Rp of 0.81 with an RMSE of 1.03 kcal/mol for the S4169 dataset as shown in Figure 2d, outstripping all existing models (Table 3). It is noteworthy that our GGL-PPI1 model, which solely relies on geometric graph-based features, demonstrated comparable performance to GGL-PPI2, outperforming TopNetTree and mCSM-PPI2 with an Rp of 0.80 and an RMSE of 1.06 kcal/mol. In the case of the S8338 dataset, we applied a stratified cross-validation approach similar to mCSM-PPI2. We ensured that hypothetical reverse mutations were consistently placed either in the training or test sets during the dataset splits, maintaining their relationship to the corresponding original mutations intact throughout the cross-validation process. GGL-PPI2 achieved an Rp of 0.85 with an RMSE of 1.07 kcal/mol as depicted in Figure 2e, and GGL-PPI1 closely followed, attaining an Rp of 0.84 with the same RMSE value. As Table 3 attests, our GGL-PPI2 is on par with TopNetTree and outperforms mCSM-PPI2 on the S8338 dataset. Table 3: Performance comparison of different methods in terms of Pearson correlation coefficients (Rp) for the single-point mutations in the SKEMPI 2.0 (S4169 and S8338) dataset. 2.2 Evaluation To evaluate our proposed model for predicting binding free energy (BFE) changes of protein-protein interactions, we consider two datasets sourced from the ProTherm database [22]. The first dataset, carefully selected by Pucci et al. [36], named S[sym] dataset. This data assembles 684 mutations from the ProTherm, comprising 342 direct mutations and their corresponding reverse mutations, resulting in a balanced dataset. The dataset specifically focuses on mutations in fifteen protein chains with solved 3D structures, ensuring highresolution data with a resolution of at least 2.5˚A. By providing experimentally measured ∆∆G values and a balanced representation of stabilizing and destabilizing mutations, the S[sym] dataset serves as a valuable resource for evaluating prediction biases in the context of predicting mutation-induced binding affinity changes. To address the issue of data leakage and enhance the generalization capability of our method, we employed the Q1744 dataset [47]. Quan et al. [48] compiled the Q3421 dataset from ProTherm, consisting of 3421 single-point mutations across 150 proteins with available PDB structures. However, the presence of homologous proteins in both the training and test set can lead to interdependent effects of mutations, compromising the model’s performance. To mitigate this, Li et al. [47] created the Q1744 dataset, derived by excluding overlapping data points and refining protein-level homology between Q3421 and S[sym] datasets, resulting in 1744 distinct mutations. Furthermore, the Q3488 dataset was created by augmenting reverse mutations in the Q1744 set. We utilized the Q3488 dataset as our training set, thereby enhancing our ∆∆G predictor’s capability to accurately predict BFE changes in PPIs. We conduct an evaluation of our model on the blind test set S[sym], with a distinct focus on both direct and reverse mutations. To assess the performance, we utilize the Pearson correlation coefficient and root-mean-square error as our primary metrics. Additionally, to discern any prediction bias, we incorporated two statistical measures: Rpdir−rev and δ. The former calculates the Pearson correlation between predictions for direct and reverse mutations, while the latter represents the sum of predicted ∆∆G values for both types of mutations. The hypothesis is that an unbiased predictor would yield Rpdir−rev = −1 and an average δ ( ¯δ) of 0 kcal/mol. Our main focus is to highlight the effectiveness of our model, GGL-PPI2, particularly emphasizing its robust geometric graph-based molecular featurization. GGL-PPI2 has demonstrated exceptional prediction accuracy, maintaining consistency for both direct and reverse mutations. As depicted in Figure 3a and 3b, our model achieves consistent Rp values of 0.57 and an RMSE of 1.28 kcal/mol, indicating its efficiency against overfitting to direct mutations. Additionally, the analysis reveals that a significant proportion of mutations fall within a prediction error of 0.5 kcal/mol and 1.0 kcal/mol, with 34.6% and 65.8% for direct mutations and 35.1% and 66.0% for reverse mutations, as depicted in Figure 3d and 3e. Furthermore, Figure 3c demonstrates that GGL-PPI2 effectively addresses prediction bias by achieving a nearly perfect Rpdir−rev value of -0.999 and an extremely low average ¯δ of 0.006 kcal/mol. Finally, the distribution plot in Figure 3f illustrates that 99.4% of mutations exhibit a prediction bias under 0.05 kcal/mol. In Table 4, we present the prediction results of our models and conduct a comprehensive comparison with other ∆∆G predictors. We observe that our GGL-PPI2 model outperforms ThermoNet [47], which was also trained on the homologyreduced set Q3488, across all evaluation measures. It outperforms ThermoNet by 21.3% for direct mutations and 18.7% for reverse mutations. Furthermore, the GGL-PPI1 model, which only uses geometric graph-based features, also performs better than ThermoNet in both direct and reverse prediction tasks. This further emphasizes the effectiveness of our geometric-graph approach. For a broader comparison against other ∆∆G predictors, we introduce the GGL-PPI2∗ model, trained on the Q6428 set constructed before the homology reduction of the set Q3421 [47]. As illustrated in Table 4, GGL-PPI2∗ excels over other methods in reverse mutation predictions. It is noteworthy that while some methods surpass GGL-PPI2∗ for direct mutations, they frequently exhibit significant bias towards reverse mutations. This paper is available on under CC 4.0 license. Arxiv

Our Datasets and Results From Our Study: GGL-PPI Models

আমাদের গবেষণা থেকে আমাদের ডেটাসেট এবং ফলাফল: GGL-PPI মডেল

Unsere Datensätze und Ergebnisse unserer Studie: GGL-PPI-Modelle

Nuestros conjuntos de datos y resultados de nuestro estudio: modelos GGL-PPI

Nos ensembles de données et résultats de notre étude : modèles GGL-PPI

हमारे डेटासेट और हमारे अध्ययन के परिणाम: जीजीएल-पीपीआई मॉडल

私たちのデータセットと研究結果: GGL-PPI モデル

연구 데이터 세트 및 결과: GGL-PPI 모델

Nossos conjuntos de dados e resultados de nosso estudo: modelos GGL-PPI

Наши наборы данных и результаты нашего исследования: модели GGL-PPI

Veri Setlerimiz ve Çalışmamızın Sonuçları: GGL-PPI Modelleri

Bộ dữ liệu và kết quả từ nghiên cứu của chúng tôi: Mô hình GGL-PPI

我们的数据集和研究结果：GGL-PPI 模型

Steam, Its Generation and Use by Babcock & Wilcox Company, is part of the HackerNoon Books Series. You can jump to any chapter in this book . BRICKWORK BOILER SETTINGS here BRICKWORK BOILER SETTINGS A consideration of the losses in boiler efficiency, due to the effects of excess air, clearly indicates the necessity of maintaining the brick setting of a boiler tight and free from air leaks. In view of the temperatures to which certain portions of such a setting are subjected, the material to be used in its construction must be of the best procurable. Boiler settings to-day consist almost universally of brickwork—two kinds being used, namely, red brick and fire brick. The red brick should only be used in such portions of the setting as are well protected from the heat. In such location, their service is not so severe as that of fire brick and ordinarily, if such red brick are sound, hard, well burned and uniform, they will serve their purpose. The fire brick should be selected with the greatest care, as it is this portion of the setting that has to endure the high temperatures now developed in boiler practice. To a great extent, the life of a boiler setting is dependent upon the quality of the fire brick used and the care exercised in its laying. The best fire brick are manufactured from the fire clays of Pennsylvania. South and west from this locality the quality of fire clay becomes poorer as the distance increases, some of the southern fire clays containing a considerable percentage of iron oxide. Until very recently, the important characteristic on which to base a judgment of the suitability of fire brick for use in connection with boiler settings has been considered the melting point, or the temperature at which the brick will liquify and run. Experience has shown, however, that this point is only important within certain limits and that the real basis on which to judge material of this description is, from the boiler man’s standpoint, the quality of plasticity under a given load. This tendency of a brick to become plastic occurs at a temperature much below the melting point and to a degree that may cause the brick to become deformed under the stress to which it is subjected. The allowable plastic or softening temperature will naturally be relative and dependent upon the stress to be endured. With the plasticity the determining factor, the perfect fire brick is one whose critical point of plasticity lies well above the working temperature of the fire. It is probable that there are but few brick on the market which would not show, if tested, this critical temperature at the stress met with in arch construction at a point less than 2400 degrees. The fact that an arch will stand for a long period under furnace temperatures considerably above this point is due entirely to the fact that its temperature as a whole is far below the furnace temperature and only about 10 per cent of its cross section nearest the fire approaches the furnace temperature. This is borne out by the fact that arches which are heated on both sides to the full temperature of an ordinary furnace will first bow down in the middle and eventually fall. A method of testing brick for this characteristic is given in the Technologic Paper No. 7 of the Bureau of Standards dealing with “The testing of clay refractories with special reference to their load carrying capacity at furnace temperatures.” Referring to the test for this specific characteristic, this publication recommends the following: “When subjected to the load test in a manner substantially as described in this bulletin, at 1350 degrees centigrade (2462 degrees Fahrenheit), and under a load of 50 pounds per square inch, a standard fire brick tested on end should show no serious deformation and should not be compressed more than one inch, referred to the standard length of nine inches.” In the Bureau of Standards test for softening temperature, or critical temperature of plasticity under the specified load, the brick are tested on end. In testing fire brick for boiler purposes such a method might be criticised, because such a test is a compression test and subject to errors from unequal bearing surfaces causing shear. Furthermore, a series of samples, presumably duplicates, will not fail in the same way, due to the mechanical variation in the manufacture of the brick. Arches that fail through plasticity show that the tensile strength of the brick is important, this being evidenced by the fact that the bottom of a wedge brick in an arch that has failed is usually found to be wider than the top and the adjacent bricks are firmly cemented together. A better method of testing is that of testing the brick as a beam subjected to its own weight and not on end. This method has been used for years in Germany and is recommended by the highest authorities in ceramics. It takes into account the failure by tension in the brick as well as by compression and thus covers the tension element which is important in arch construction. The plastic point under a unit stress of 100 pounds per square inch, which may be taken as the average maximum arch stress, should be above 2800 degrees to give perfect results and should be above 2400 degrees to enable the brick to be used with any degree of satisfaction. The other characteristics by which the quality of a fire brick is to be judged are: Fusion point. In view of the fact that the critical temperature of plasticity is below the fusion point, this is only important as an indication from high fusion point of a high temperature of plasticity. Hardness. This is a relative quality based on an arbitrary scale of 10 and is an indication of probable cracking and spalling. Expansion. The lineal expansion per brick in inches. This characteristic in conjunction with hardness is a measure of the physical movement of the brick as affecting a mass of brickwork, such movement resulting in cracked walls, etc. The expansion will vary between wide limits in different brick and provided such expansion is not in excess of, say, .05 inch in a 9-inch brick, when measured at 2600 degrees, it is not particularly important in a properly designed furnace, though in general the smaller the expansion the better. Compression. The strength necessary to cause crushing of the brick at the center of the 4½ inch face by a steel block one inch square. The compression should ordinarily be low, a suggested standard being that a brick show signs of crushing at 7500 pounds. Size of Nodules. The average size of flint grains when the brick is carefully crushed. The scale of these sizes may be considered: Small, size of anthracite rice; large, size of anthracite pea. Ratio of Nodules. The percentage of a given volume occupied by the flint grains. This scale may be considered: High, 90 to 100 per cent; medium, 50 to 90 per cent; low, 10 to 50 per cent. The statement of characteristics suggested as desirable, are for arch purposes where the hardest service is met. For side wall purposes the compression and hardness limit may be raised considerably and the plastic point lowered. Aside from the physical properties by which a fire brick is judged, it is sometimes customary to require a chemical analysis of the brick. Such an analysis is only necessary as determining the amount of total basic fluxes (K2O, Na2O, CaO, MgO and FeO). These fluxes are ordinarily combined into one expression, indicated by the symbol RO. This total becomes important only above 0.2 molecular equivalent as expressed in ceramic empirical formulae, and this limit should not be exceeded. [75] From the nature of fire brick, their value can only be considered from a relative standpoint. Generally speaking, what are known as first-grade fire brick may be divided into three classes, suitable for various conditions of operation, as follows: Class A. For stoker-fired furnaces where high overloads are to be expected or where other extreme conditions of service are apt to occur. Class B. For ordinary stoker settings where there will be no excessive overloads required from the boiler or any hand-fired furnaces where the rates of driving will be high for such practice. Class C. For ordinary hand-fired settings where the presumption is that the boilers will not be overloaded except at rare intervals and for short periods only. gives the characteristics of these three classes according to the features determining the quality. indicates that the hardness of the brick in general increases with the poorer qualities. Provided the hardness is sufficient to enable the brick to withstand its load, additional hardness is a detriment rather than an advantage. Table 61 This table An approximate determination of the quality of a fire brick may be made from the appearance of a fracture. Where such a fracture is open, clean, white and flinty, the brick in all probability is of a good quality. If this fracture has the fine uniform texture of bread, the brick is probably poor. In considering the heavy duty of brick in boiler furnaces, experience shows that arches are the only part that ordinarily give trouble. These fail from the following causes: Bad workmanship in laying up of brick. This feature is treated below. The tendency of a brick to become plastic at a temperature below the fusing point. The limits of allowable plastic temperature have already been pointed out. Spalling. This action occurs on the inner ends of combustion arches where they are swept by gases at a high velocity at the full furnace temperature. The most troublesome spalling arises through cold air striking the heated brickwork. Failure from this cause is becoming rare, due to the large increase in number of stoker installations in which rapid temperature changes are to a great degree eliminated. Furthermore, there are a number of brick on the market practically free from such defects and where a new brick is considered, it can be tried out and if the defect exists, can be readily detected and the brick discarded. Failures of arches from the expansive power of brick are also rare, due to the fact that there are a number of brick in which the expansion is well within the allowable limits and the ease with which such defects may be determined before a brick is used. Failures through chemical disintegration. Failure through this cause is found only occasionally in brick containing a high percentage of iron oxide. With the grade of brick selected best suited to the service of the boiler to be set, the other factor affecting the life of the setting is the laying. It is probable that more setting difficulties arise from the improper workmanship in the laying up of brick than from poor material, and to insure a setting which will remain tight it is necessary that the masonry work be done most carefully. This is particularly true where the boiler is of such a type as to require combustion arches in the furnace. Red brick should be laid in a thoroughly mixed mortar composed of one volume of Portland cement, 3 volumes of unslacked lime and 16 volumes of clear sharp sand. Not less than 2½ bushels of lime should be used in the laying up of 1000 brick. Each brick should be thoroughly embedded and all joints filled. Where red brick and fire brick are both used in the same wall, they should be carried up at the same time and thoroughly bonded to each other. All fire brick should be dry when used and protected from moisture until used. Each brick should be dipped in a thin fire clay wash, “rubbed and shoved” into place, and tapped with a wooden mallet until it touches the brick next below it. It must be recognized that fire clay is not a cement and that it has little or no holding power. Its action is that of a filler rather than a binder and no fire-clay wash should be used which has a consistency sufficient to permit the use of a trowel. All fire-brick linings should be laid up four courses of headers and one stretcher. Furnace center walls should be entirely of fire brick. If the center of such walls are built of red brick, they will melt down and cause the failure of the wall as a whole. Fire-brick arches should be constructed of selected brick which are smooth, straight and uniform. The frames on which such arches are built, called arch centers, should be constructed of batten strips not over 2 inches wide. The brick should be laid on these centers in courses, not in rings, each joint being broken with a bond equal to the length of half a brick. Each course should be first tried in place dry, and checked with a straight edge to insure a uniform thickness of joint between courses. Each brick should be dipped on one side and two edges only and tapped into place with a mallet. Wedge brick courses should be used only where necessary to keep the bottom faces of the straight brick course in even contact with the centers. When such contact cannot be exactly secured by the use of wedge brick, the straight brick should lean away from the center of the arch rather than toward it. When the arch is approximately two-thirds completed, a trial ring should be laid to determine whether the key course will fit. When some cutting is necessary to secure such a fit, it should be done on the two adjacent courses on the side of the brick away from the key. It is necessary that the keying course be a true fit from top to bottom, and after it has been dipped and driven it should not extend below the surface of the arch, but preferably should have its lower ledge one-quarter inch above this surface. After fitting, the keys should be dipped, replaced loosely, and the whole course driven uniformly into place by means of a heavy hammer and a piece of wood extending the full length of the keying course. Such a driving in of this course should raise the arch as a whole from the center. The center should be so constructed that it may be dropped free of the arch when the key course is in place and removed from the furnace without being burned out. Care of Brickwork—Before a boiler is placed in service, it is essential that the brickwork setting be thoroughly and properly dried, or otherwise the setting will invariably crack. The best method of starting such a process is to block open the boiler damper and the ashpit doors as soon as the brickwork is completed and in this way maintain a free circulation of air through the setting. If possible, such preliminary drying should be continued for several days before any fire is placed in the furnace. When ready for the drying out fire, wood should be used at the start in a light fire which may be gradually built up as the walls become warm. After the walls have become thoroughly heated, coal may be fired and the boiler placed in service. As already stated, the life of a boiler setting is dependent to a large extent upon the material entering into its construction and the care with which such material is laid. A third and equally important factor in the determining of such life is the care given to the maintaining of the setting in good condition after the boiler is placed in operation. This feature is discussed more fully in the chapter dealing with general boiler room management. Steel Casings—In the chapter dealing with the losses operating against high efficiencies as indicated by the heat balance, it has been shown that a considerable portion of such losses is due to radiation and to air infiltration into the boiler setting. These losses have been variously estimated from 2 to 10 per cent, depending upon the condition of the setting and the amount of radiation surface, the latter in turn being dependent upon the size of the boiler used. In the modern efforts after the highest obtainable plant efficiencies much has been done to reduce such losses by the use of an insulated steel casing covering the brickwork. In an average size boiler unit the use of such casing, when properly installed, will reduce radiation losses from one to two per cent., over what can be accomplished with the best brick setting without such casing and, in addition, prevent the loss due to the infiltration of air, which may amount to an additional five per cent., as compared with brick settings that are not maintained in good order. Steel plate, or steel plate backed by asbestos mill-board, while acting as a preventative against the infiltration of air through the boiler setting, is not as effective from the standpoint of decreasing radiation losses as a casing properly insulated from the brick portion of the setting by magnesia block and asbestos mill-board. A casing which has been found to give excellent results in eliminating air leakage and in the reduction of radiation losses is clearly illustrated on page . 306 Many attempts have been made to use some material other than brick for boiler settings but up to the present nothing has been found that may be considered successful or which will give as satisfactory service under severe conditions as properly laid brickwork. About HackerNoon Book Series: We bring you the most important technical, scientific, and insightful public domain books. This book is part of the public domain. Babcock & Wilcox Company (2007). Steam, Its Generation and Use. Urbana, Illinois: Project Gutenberg. Retrieved https://www.gutenberg.org/cache/epub/22657/pg22657-images.html This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at , located at . www.gutenberg.org https://www.gutenberg.org/policy/license.html

BRICKWORK BOILER SETTINGS

On Combinations of Masters against the public

Steam, Its Generation and Use by Babcock & Wilcox Company, is part of the HackerNoon Books Series. You can jump to any chapter in this book . THE BABCOCK & WILCOX BOILER here THE BABCOCK & WILCOX BOILER The following brief description of the Babcock & Wilcox boiler will clearly indicate the manner in which it fulfills the requirements of the perfect steam boiler already enumerated. The Babcock & Wilcox boiler is built in two general classes, the longitudinal drum type and the cross drum type. Either of these designs may be constructed with vertical or inclined headers, and the headers in turn may be of wrought steel or cast iron dependent upon the working pressure for which the boiler is constructed. The headers may be of different lengths, that is, may connect different numbers of tubes, and it is by a change in the number of tubes in height per section and the number of sections in width that the size of the boiler is varied. The longitudinal drum boiler is the generally accepted standard of Babcock & Wilcox construction. The cross drum boiler, though originally designed to meet certain conditions of headroom, has become popular for numerous classes of work where low headroom is not a requirement which must be met. LONGITUDINAL DRUM CONSTRUCTION—The heating surface of this type of boiler is made up of a drum or drums, depending upon the width of the boiler extending longitudinally over the other pressure parts. To the drum or drums there are connected through cross boxes at either end the sections, which are made up of headers and tubes. At the lower end of the sections there is a mud drum extending entirely across the setting and connected to all sections. The connections between all parts are by short lengths of tubes expanded into bored seats. The drums are of three sheets, of such thickness as to give the required factor of safety under the maximum pressure for which the boiler is constructed. The circular seams are ordinarily single lap riveted though these may be double lap riveted to meet certain requirements of pressure or of specifications. The longitudinal seams are properly proportioned butt and strap or lap riveted joints dependent upon the pressure for which the boilers are built. Where butt strap joints are used the straps are bent to the proper radius in an hydraulic press. The courses are built independently to template and are assembled by an hydraulic forcing press. All riveted holes are punched one-quarter inch smaller than the size of rivets as driven and are reamed to full size after the plates are assembled. All rivets are driven by hydraulic pressure and held until black. The drumheads are hydraulic forged at a single heat, the manhole opening and stiffening ring being forged in position. Flat raised seats for water column and feed connections are formed in the forging. All heads are provided with manholes, the edges of which are turned true. The manhole plates are of forged steel and turned to fit manhole opening. These plates are held in position by forged-steel guards and bolts. The drum nozzles are of forged steel, faced, and fitted with taper thread stud bolts. Cross boxes by means of which the sections are attached to the drums, are of forged steel, made from a single sheet. Where two or more drums are used in one boiler they are connected by a cross pipe having a flanged outlet for the steam connection. The sections are built of 4-inch hot finished seamless open-hearth steel tubes of No. 10 B. W. G. where the boilers are built for working pressures up to 210 pounds. Where the working pressure is to be above this and below 260 pounds, No. 9 B. W. G. tubes are supplied. The tubes are expanded into headers of serpentine or sinuous form, which dispose the tubes in a staggered position when assembled as a complete boiler. These headers are of wrought steel or of cast iron, the latter being ordinarily supplied where the working pressure is not to exceed 160 pounds. The headers may be either vertical or inclined as shown in the various illustrations of assembled boilers. Opposite each tube end in the headers there is placed a handhole of sufficient size to permit the cleaning, removal or renewal of a tube. These openings in the wrought steel vertical headers are elliptical in shape, machine faced, and milled to a true plane back from the edge a sufficient distance to make a seat. The openings are closed by inside fitting forged plates, shouldered to center in the opening, their flanged seats milled to a true plane. These plates are held in position by studs and forged-steel binders and nuts. The joints between plates and headers are made with a thin gasket. In the wrought-steel inclined headers the handhole openings are either circular or elliptical, the former being ordinarily supplied. The circular openings have a raised seat milled to a true plane. The openings are closed on the outside by forged-steel caps, milled and ground true, held in position by forged-steel safety clamps and secured by ball-headed bolts to assure correct alignment. With this style of fitting, joints are made tight, metal to metal, without packing of any kind. Where elliptical handholes are furnished they are faced inside, closed by inside fitting forged-steel plates, held to their seats by studs and secured by forged-steel binders and nuts. The joints between plates and header are made with a thin gasket. The vertical cast-iron headers have elliptical handholes with raised seats milled to a true plane. These are closed on the outside by cast-iron caps milled true, held in position by forged-steel safety clamps, which close the openings from the inside and which are secured by ball-headed bolts to assure proper alignment. All joints are made tight, metal to metal, without packing of any kind. The mud drum to which the sections are attached at the lower end of the rear headers, is a forged-steel box 7¼ inches square, and of such length as to be connected to all headers by means of wrought nipples expanded into counterbored seats. The mud drum is furnished with handholes for cleaning, these being closed from the inside by forged-steel plates with studs, and secured on a faced seat in the mud drum by forged-steel binders and nuts. The joints between the plates and the drum are made with thin gaskets. The mud drum is tapped for blow-off connection. All connections between drums and sections and between sections and mud drum are of hot finished seamless open-hearth steel tubes of No. 9 B. W. G. Boilers of the longitudinal drum type are suspended front and rear from wrought-steel supporting frames entirely independent of the brickwork. This allows for expansion and contraction of the pressure parts without straining either the boiler or the brickwork, and also allows of brickwork repair or renewal without in any way disturbing the boiler or its connections. CROSS DRUM CONSTRUCTION—The cross drum type of boilers differs from the longitudinal only in drum construction and method of support. The drum in this type is placed transversely across the rear of the boiler and is connected to the sections by means of circulating tubes expanded into bored seats. The drums for all pressures are of two sheets of sufficient thickness to give the required factor of safety. The longitudinal seams are double riveted butt strapped, the straps being bent to the proper radius in an hydraulic press. The circulating tubes are expanded into the drums at the seams, the butt straps serving as tube seats. The drumheads, drum fittings and features of riveting are the same in the cross drum as in the longitudinal types. The sections and mud drum are also the same for the two types. Cross drum boilers are supported at the rear on the mud drum which rests on cast-iron foundation plates. They are suspended at the front from a wrought-iron supporting frame, each section being suspended independently from the cross members by hook suspension bolts. This method of support is such as to allow for expansion and contraction without straining either the boiler or the brickwork and permits of repair or renewal of the latter without in any way disturbing the boiler or its connections. The following features of design and of attachments supplied are the same for all types. FRONTS—Ornamental fronts are fitted to the front supporting frame. These have large doors for access to the front headers and panels above the fire fronts. The fire fronts where furnished have independent frames for fire doors which are bolted on, and ashpit doors fitted with blast catches. The lugs on door frames and on doors are cast solid. The faces of doors and of frames are planed and the lugs milled. The doors and frames are placed in their final relative position, clamped, and the holes for hinge pins drilled while thus held. A perfect alignment of door and frame is thus assured and the method is representative of the care taken in small details of manufacture. The front as a whole is so arranged that any stoker may be applied with but slight modification wherever boilers are set with sufficient furnace height. In the vertical header boilers large wrought-iron doors, which give access to the rear headers, are attached to the rear supporting frame. FITTINGS—Each boiler is provided with the following fittings as part of the standard equipment: Blow-off connections and valves attached to the mud drum. Safety valves placed on nozzles on the steam drums. A water column connected to the front of the drum. A steam gauge attached to the boiler front. Feed water connection and valves. A flanged stop and check valve of heavy pattern is attached directly to each drumhead, closing automatically in case of a rupture in the feed line. All valves and fittings are substantially built and are of designs which by their successful service for many years have become standard with The Babcock & Wilcox Co. The fixtures that are supplied with the boilers consist of: Dead plates and supports, the plates arranged for a fire brick lining. A full set of grate bars and bearers, the latter fitted with expansion sockets for side walls. Flame bridge plates with necessary fastenings, and special fire brick for lining same. Bridge wall girder for hanging bridge wall with expansion sockets for side walls. A full set of access and cleaning doors through which all portions of the pressure parts may be reached. A swing damper and frame with damper operating rig. There are also supplied with each boiler a wrench for handhole nuts, a water-driven turbine tube cleaner, a set of fire tools and a metal steam hose and cleaning pipe equipped with a special nozzle for blowing dust and soot from the tubes. Aside from the details of design and construction as covered in the foregoing description, a study of the illustrations will make clear the features of the boiler as a whole which have led to its success. The method of supporting the boiler has been described. This allows it to be hung at any height that may be necessary to properly handle the fuel to be burned or to accommodate the stoker to be installed. The height of the nest of tubes which forms the roof of the furnace is thus the controlling feature in determining the furnace height, or the distance from the front headers to the floor line. The sides and front of the furnace are formed by the side and front boiler walls. The rear wall of the furnace consists of a bridge wall built from the bottom of the ashpit to the lower row of tubes. The location of this wall may be adjusted within limits to give the depth of furnace demanded by the fuel used. Ordinarily the bridge wall is the determining feature in the locating of the front baffle. Where a great depth of furnace is necessary, in which case, if the front baffle were placed at the bridge wall the front pass of the boiler would be relatively too long, a patented construction is used which maintains the baffle in what may be considered its normal position, and a connection made between the baffle and the bridge wall by means of a tile roof. Such furnace construction is known as a “Webster” furnace. A consideration of this furnace will clearly indicate its adaptability, by reason of its flexibility, for any fuel and any design of stoker. The boiler lends itself readily to installation with an extension or Dutch oven furnace if this be demanded by the fuel to be used, and in general it may be stated that a satisfactory furnace arrangement may be made in connection with a Babcock & Wilcox boiler for burning any fuel, solid, liquid or gaseous. The gases of combustion evolved in the furnace above described are led over the heating surfaces by two baffles. These are formed of cast-iron baffle plates lined with special fire brick and held in position by tube clamps. The front baffle leads the gases through the forward portion of the tubes to a chamber beneath the drum or drums. It is in this chamber that a superheater is installed where such an apparatus is desired. The gases make a turn over the front baffle, are led downward through the central portion of the tubes, called the second pass, by means of a hanging bridge wall of brick and the second baffle, around which they make a second turn upward, pass through the rear portion of the tubes and are led to the stack or flue through a damper box in the rear wall, or around the drums to a damper box placed overhead. The space beneath the tubes between the bridge wall and the rear boiler wall forms a pocket into which much of the soot from the gases in their downward passage through the second pass will be deposited and from which it may be readily cleaned through doors furnished for the purpose. The gas passages are ample and are so proportioned that the resistance offered to the gases is only such as will assure the proper abstraction of heat from the gases without causing undue friction, requiring excessive draft. The method in which the feed water is introduced through the front drumhead of the boiler is clearly seen by reference to the illustration. From this point of introduction the water passes to the rear of the drum, downward through the rear circulating tubes to the sections, upward through the tubes of the sections to the front headers and through these headers and front circulating tubes again to the drum where such water as has not been formed into steam retraces its course. The steam formed in the passage through the tubes is liberated as the water reaches the front of the drum. The steam so formed is stored in the steam space above the water line, from which it is drawn through a so-called “dry pipe.” The dry pipe in the Babcock & Wilcox boiler is misnamed, as in reality it fulfills none of the functions ordinarily attributed to such a device. This function is usually to restrict the flow of steam from a boiler with a view to avoid priming. In the Babcock & Wilcox boiler its function is simply that of a collecting pipe, and as the aggregate area of the holes in it is greatly in excess of the area of the steam outlet from the drum, it is plain that there can be no restriction through this collecting pipe. It extends nearly the length of the drum, and draws steam evenly from the whole length of the steam space. Patented Side Dusting Doors The large tube doors through which access is had to the front headers and the doors giving such access to the rear headers in boilers of the vertical header type have already been described and are shown clearly by the illustrations on pages and . In boilers of the inclined header type, access to the rear headers is secured through the chamber formed by the headers and the rear boiler wall. Large doors in the sides of the setting give full access to all parts for inspection and for removal of accumulations of soot. Small dusting doors are supplied for the side walls through which all of the heating surfaces may be cleaned by means of a steam dusting lance. These side dusting doors are a patented feature and the shutters are self closing. In wide boilers additional cleaning doors are supplied at the top of the setting to insure ease in reaching all portions of the heating surface. 56 74 The drums are accessible for inspection through the manhole openings. The removal of the handhole plates makes possible the inspection of each tube for its full length and gives the assurance that no defect can exist that cannot be actually seen. This is particularly advantageous when inspecting for the presence of scale. The materials entering into the construction of the Babcock & Wilcox boiler are the best obtainable for the special purpose for which they are used and are subjected to rigid inspection and tests. The boilers are manufactured by means of the most modern shop equipment and appliances in the hands of an old and well-tried organization of skilled mechanics under the supervision of experienced engineers. About HackerNoon Book Series: We bring you the most important technical, scientific, and insightful public domain books. This book is part of the public domain. Babcock & Wilcox Company (2007). Steam, Its Generation and Use. Urbana, Illinois: Project Gutenberg. Retrieved https://www.gutenberg.org/cache/epub/22657/pg22657-images.html This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at , located at . www.gutenberg.org https://www.gutenberg.org/policy/license.html

THE BABCOCK & WILCOX BOILER

Scientific American, Vol. XXXVII.—No. 2. [New Series.], July 14, 1877 by Various, is part of the HackerNoon Books Series. You can jump to any chapter in this book . A NEW METHOD OF BOOKBINDING. here A NEW METHOD OF BOOKBINDING. The annexed engravings represent a new system of binding books, for which a number of important advantages are claimed. It obviates stitching, allows of each leaf being firmly secured, and hence is especially well suited for single-leaved books. It admits of plates and maps being bound in their proper places instead of being pasted in, and renders the book much stronger and more durable. The inventor claims a saving of 40 to 75 per cent of the time required for stitching, and of 50 per cent of the time needed in ordinary rebinding work. The mode of operation is as follows: On receiving the sheets, the binder folds them and places them in consecutive order, according to the printer's signature. The front and bottom edges of the book are then trimmed so as to obtain two straight sides; and the backs of the sheets are cut off, transforming them into single leaves. Horizontal lines are now marked with pencil across the back of the book for the saw cuts; and a diagonal line, A, B, Fig. 2, is drawn to serve as a guide in replacing the leaves in their proper places. A thin coat of glue is next applied to the back; and when this is dry, the book is divided into sections of from four to eight leaves (without counting them) entirely disregarding the printer's signatures, but placing the sheets in their original order. The binder places the first section removed at his right hand, the next at his left, and so on, forming two piles. Each pile is then straightened, and in the back of each, a little below the transverse lines, are made bevel cuts with the saw. Said cuts are ⅛ inch in length, inclined at an angle of 45°, and so placed that one half their length is above and the other half below the marked line. When one pile of sheets is thus sawn, the other pile is similarly treated; but the corresponding cuts are made at relatively opposite angles. This will be understood from Fig. 1, in which C represents the edge of the right hand pile, for example, and D that of the left hand pile. The sections of each pile are now returned in their regular order, according to the printer's signatures. Should a section have been misplaced, the diagonal line, being thus broken, will show the fact. It will be seen, however, that this arrangement involves the alternate use of sheets from each pile, so that, when all are put together, the beveled cuts will cross or form dovetails, as shown in Fig. 3. Half inch strips of white paper muslin, E, Fig. 4, are next pasted around the back edges of the first and last sections. This is done to strengthen the hold of the twines in the back of the book, said sections necessarily bearing the whole strain of the covers. The twine used corresponds in size to the holes made by the coincidence of the beveled saw cuts. This twine is passed through the holes by means of a blunt darning needle. The back of the book is shown in Fig. 2; and in Fig. 4 the twines are represented as passed. Nothing further remains to be done but to paste in the fly-leaves and lining, and finish the book in the usual manner. It is evident that this a very much stronger method of securing the leaves than that in which the twine is simply laid and glued in a straight cut. Each leaf is independently fastened; and the thread is prevented from cutting through, as is commonly the case when the book has been used to any great extent. Books can be bound to open more or less as desired; and in rebinding, instead of taking the book apart and cutting threads, a thin shaving is sliced off the back, and the leaves are treated in the manner already described. Patented March 20, 1877, by Mr. Florenz E. Schmitz. For further information, address Messrs. Schmitz and Slosson, box 1180, Middletown, Orange county, N. Y. About HackerNoon Book Series: We bring you the most important technical, scientific, and insightful public domain books. This book is part of the public domain. Various (2012). Scientific American, Vol. XXXVII. —No. 2. [New Series.], July 14, 1877. Urbana, Illinois: Project Gutenberg. Retrieved https://www.gutenberg.org/cache/epub/38481/pg38481-images.html This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at , located at . www.gutenberg.org https://www.gutenberg.org/policy/license.html

A NEW METHOD OF BOOKBINDING

The Romance of Modern Mechanism by Archibald Williams is part of the HackerNoon Books Series. You can jump to any chapter in this book . CHAPTER XXIV here AGRICULTURAL MACHINERY PLOUGHS — DRILLS AND SEEDERS — REAPING MACHINES — THRESHING MACHINES — PETROL-DRIVEN FIELD MACHINERY— ELECTRICAL FARMING MACHINERY Agriculture is at once the oldest and most important of all national industries. Man being a graminivorous animal—witness his molar, or grinding, "double" teeth—has, since the earliest times, been obliged to observe the seasons, planting his crops when the ground is moist, and reaping them when the weather is warm and dry. Apart from the nomad races of the deserts and steppes, who find their chief subsistence in the products of the date-palm and of their flocks and herds, all nations cultivate a large portion of the country which they inhabit. Ancient monuments, the oldest inscriptions and writings, bear witness to the prime importance of the plough and reaping-hook; and it may be reasonably assumed that the progress of civilisation is proved by the increased use of cereal foods, and better methods of garnering and preparing them. For thousands of years the sickle, which Greek and Roman artists placed in the hand of their Goddess of the Harvest, and the rude plough, consisting of, perhaps, only a crooked bough with a pointed end, were practically the only implements known to the husbandman besides his spade and mattock. Where labour is abundant and each householder has time to cultivate the little plot which suffices for the maintenance of his own family, and while there is little inducement to take part in other than agricultural industries—tedious and time-wasting methods have held their own. But in highly civilised communities carrying on manufactures of all sorts it is difficult for the farmer to secure an abundance of human help, and yet it is recognised that a speedy preparation and sowing of the land, and a prompt gathering and threshing of the harvest, is all in favour of producing a successful and well-conditioned crop. In England, eighty years ago, three men lived in the country for every one who lived in the town. Now the proportion has been reversed; and that not in the British Isles alone. The world does not mean to starve; but civilisation demands that as few people as possible should be devoted to procuring the "staff of life" for both man and beast. We should reasonably expect, therefore, that the immense advance made in mechanical science during the last century should have left a deep mark on agricultural appliances. Such an expectation is more than justified; for are there not many among us who have seen the sickle and the flail at work where now the "self-binder" and threshing machine perform the same duties in a fraction of the time formerly required? The ploughman, plodding sturdily down the furrow behind his clever team, is indeed still a common sight; but in the tilling season do we not hear the snort of the steam-engine, as its steel rope tears a six-furrow plough through the mellow earth? When the harvest comes we realise even more clearly how largely machinery has supplanted man; while in the processes of separating the grain from its straw the human element plays an even smaller part. It would not be too much to say that, were we to revert next year to the practices of our grandfathers, we should starve in the year following. This chapter will be confined to a consideration of machinery operated by horse, steam, or other power, which falls under four main headings,—ploughs, drills, reapers, and threshers. PLOUGHS The firm of Messrs. John Fowler and Company, of Leeds, is most intimately connected with the introduction of the steam plough and cultivator. Their first type of outfit included one engine only, the traversing of the plough across the field being effected by means of cables passing round a pulley on a low, four-wheeled truck, moved along the opposite edge of the field by ropes dragging on an anchor. Another method was to have the engine stationary at one corner of the field, and an anchor at each of the three other corners, the two at the ends of the furrow being moved for every journey of the plough. In, or about, the year 1865 this arrangement succumbed to the simple and, as it now seems to us, obvious improvement of introducing a second engine to progress vis-à-vis with the first, and do its share of the pulling. The modern eight-furrow steam plough will turn ten acres a day quite easily, at a much lower cost than that of horse labour. For tearing up land after a crop "cultivators" are sometimes used. They have arrowhead-shaped [321]coulters, which cut very deep and bring large quantities of fresh earth to the surface. The ground is now pulverised by harrows of various shapes, according to the nature of the crop to be sown. English farmers generally employ the spike harrow; but Yankee agriculturists make great use of the spring-tooth form, which may best be described as an arrangement of very strong springs much resembling in outline the springs of house bells. The shorter arm is attached to the frame, while the longer and pointed arm tears the earth. DRILLS AND SEEDERS In highly civilised countries the man carrying a basket from which he flings seeds broadcast is a very rare sight indeed. The primitive method may have been effective—a good sower could cover an acre evenly with half a pint of turnip seed—but very slow. We now use a long bin mounted on wheels, which revolves discs inside the bin, furnished with tiny spoons round the periphery to scoop small quantities of seed into tubes terminating in a coulter. The farmer is thus certain of having evenly planted and parallel rows of grain, which in the early spring, when the sprouting begins, make so pleasant an addition to the landscape. The "corn," or maize, crop of the United States is so important that it demands special sowing machinery, which plants single grains at intervals of about eighteen inches. A somewhat similar device is used for planting potatoes. Passing over the weeding machines, which offer no features of particular interest, we come to the REAPING MACHINES, on which a vast amount of ingenuity has been expended. At the beginning of the nineteenth century the Royal Agricultural Society of Great Britain offered a prize for the introduction of a really useful machine which should replace the scythe and sickle. Several machines were brought out, but they did not prove practical enough to attract much attention. Cyrus H. McCormick invented in 1831 the reaper, which, with very many improvements added, is to-day employed in all parts of the world. The most noticeable point of this machine was the bar furnished with a row of triangular blades which passed very rapidly to and fro through slots in an equal number of sharp steel points, against which they cut the grain. The to-and-fro action of the cutter-blade was produced by a connecting-rod working on a crank rotated by the wheels carrying the machine. The first McCormick reaper did wonders on a Virginian farm; other inventors were stimulated; and in 1833 there appeared the Hussey reaper, built on somewhat similar lines. For twelve years or so these two machines competed against one another all over the United States; and then McCormick added a raker attachment, which, when sufficient grain had accumulated on the platform, enabled a second man on the machine to sweep it off to be tied up into a sheaf. At the Great Exhibition held in London in 1851, the judges awarded a special medal to the inventor, reporting that the whole expense of the Exhibition would have been well recouped if only the reaper were introduced into England. From France McCormick received the decoration of the Legion of Honour "for having done more for the cause of agriculture than any man then living." It would be reasonable to expect that, after this public recognition, the mechanical reaper would have been immediately valued at its true worth. "Yet no man had more difficulty in introducing his machines than that pioneer inventor of agricultural implements. Farmers everywhere were slow to accept it, and manufacturers were unwilling to undertake its manufacture. Even after the value of the machine had been demonstrated, everyone seemed to fear that it would break down on rocky and uneven fields; and the inventor had to demonstrate in person to the farmers the practicability of the reapers, and then even guarantee them before the money could be obtained. Through all these trying discouragements the persistent inventor passed before he saw any reward for the work that he had spent half a lifetime in perfecting. The ultimate triumph of the inventor may be sufficient reward for his labours and discouragements, but those who would begrudge him the wealth that he subsequently made from his invention should consider some of the difficulties and obstacles he had to overcome in the beginning." [24] In 1858 an attachment was fitted to replace the second passenger on the machine. Four men followed behind to tie up the grain as it was shot off the machine. Inventors tried to abolish the need for these extra hands by means of a self-binding device. A practical method, employing wire, appeared in 1860; but so great was the trouble caused by stray pieces of the wire getting into threshing and other machinery through which the grain subsequently passed that farmers went back to hand work, until the Appleby patent of 1873 replaced wire by twine. Words alone would convey little idea of how the corn is collected and encircled with twine; how the knot is tied by an ingenious shuttle mechanism; and how it is thrown out into a set of arms which collect sufficient sheaves to form a "stook" before it lets them fall. So we would advise our readers to take the next chance of examining a modern self-binder, and to persuade the man in charge to give as lucid an explanation as he can of the way in which things are done. Popular prejudice having once been conquered, the success of the reapers was assured. The year 1870 saw 60,000 in use; by 1885 the output had increased to 250,000; and to-day the manufacture of agricultural labour-saving machines gives employment to over 200,000 people; an equal number being occupied in their transport and sale in all parts of the globe. In California, perhaps more than in any other country, "power" agricultural machinery is seen at its best. Great traction-engines here take the place of human labour to an extraordinary extent. The largest, of 50 h.p. and upwards, "with driving-wheels 60 inches in diameter and flanges of generous width, travel over the uneven surface of the grain fields, crossing ditches and low places, and ascending the sides of steep hills, with as much apparent ease as a locomotive rolls along its steel rails. Such powerful traction-engines, or 'automobiles' as they are commonly called by the American farmers, are capable of dragging behind them sixteen 10-inch ploughs, four 6-foot harrows, and a drill and seeder. The land is thus ploughed, drilled, and seeded all at one time. From fifty to seventy-five acres of virgin soil can thus be ploughed and planted in a single day. When the harvest comes the engines are again brought into service, and the fields that would ordinarily defy the best efforts of an army of workmen are garnered quickly and easily. The giant harvester is hitched to the traction-engine in place of the ploughs and harrows, and cuts, binds, and stacks the golden wheat from seventy-five acres in a single day. The cutters are 26 feet wide, and they make a clear swathe across the field. Some of them thresh, clean, and sack the wheat as fast as it is cut and bound. Other traction-engines follow to gather up the sacked wheat, and whole train-loads of it thus move across the field to the granaries or railways of the seaboard or interior." For "dead ripe" crops the "header" is often used in California. Instead of being pulled it is by mules, and merely cuts off the heads, leaving the straw to be trampled down by the animals since it has no value. Swathes as wide as 50 feet are thus treated, the grain being threshed out while the machine moves. pushed One of the most beautiful, and at the same time useful, crops in the world is that of maize, which feeds not only vast numbers of human beings, but also countless flocks and herds, the latter eating the green stalks as well as the ripened grain. The United States alone produced no less than 2,523,648,312 bushels of this cereal in 1902, as against 987,000,000 bushels of wheat, and 670,000,000 bushels of barley. Now, maize has a very tough stalk, often 10 feet high and an inch thick, which cannot be cut with the ease of wheat or barley. So a special machine has been devised to handle it. The row of corn is picked up, if fallen, by chains furnished with projecting spikes working at an angle to the perpendicular, so as to lift and simultaneously pull back the stalks, which pass into a horizontal V-shaped frame. This has a broad opening in front, but narrows towards its rear end, where stationary sickles fixed on either side give the stalk a drawing cut before it reaches the single knife moving to right and left in the angle of the V, which severs the stalk completely. The McCormick machine gathers the corn in vertical bundles, and ties them up ready for the "shockers." THRESHING MACHINES In principle these are simple enough. The straw and grain is fed into a slot and pulled down between a toothed rotating drum and a fixed toothed concave. These tear out the grain from the ear. The former falls into the hopper of a winnowing and riddling machine, which clears it from dust and husks, and allows it to pass to a hopper. An endless chain of buckets carries it to the delivery bins, holding just one sackful each, which when full discharge the grain through spouts into the receptacles waiting below their mouths. An automatic counter records the number of sackfuls of corn that have been discharged, so that dishonesty on the part of employés becomes practically an impossibility. While the grain is thus treated, oscillating rakes have arranged the straw and shaken it out behind in a form convenient for binding, and the chaff has passed to its proper heap, to be used as fuel for the engine or as food for cattle. PETROL-DRIVEN FIELD MACHINERY On water, rail, and road the petrol engine has entered into rivalry with steam—very successfully too. And now it bids fair to challenge both steam-engine and horse as the motive power for agricultural operations. Probably the best-known English petrol-driven farmer's help is that made by Mr. Dan Albone, of Biggleswade, who in past times did much to introduce the safety bicycle to the public. The "Ivel" motor is not beautiful to look upon; its sides are slab, its outlines rather suggestive of an inverted punt. But it is a willing and powerful worker; requires no feeding in the early hours of the morning; no careful brush down after the day's work; no halts to ease wearied muscles. In one tank is petrol, in another lubricating oil, in a third water to keep the cylinders cool. A double-cylinder motor of 18 h.p. transmits its energy through a large clutch and train of cogs to the road wheels, made extra wide and well corrugated so that they shall not sink into soft ground or slip on hard. There is a broad pulley-wheel peeping out from one side of the machine, which is ready to drive chaff-cutters or threshers, pump, grind corn, or turn a dynamo at a moment's notice. Hitch the "Ivel" on to a couple of reapers or a three-furrow plough, and it soon shows its superiority to "man's friend." Here are some records:— We may quote a paragraph from the pen of "Home Counties," a well-known and perspicacious writer on agricultural topics. "It is because motor-farming is likely to result in a more thorough cultivation of the land and a more skilful and more enlightened practice of agriculture, and not in a further extension of those deplorable land-scratching and acre-grasping methods of which so many pitiful examples may be seen on our clay soils, that its beginnings are being sympathetically watched by many people who have the best interests of the rural districts and the prosperity of agriculture at heart." [25] Will our farmers give the same welcome to the agricultural motor that was formerly accorded to the mechanical reaper? Prophecy is risky, but if, before a decade has elapsed, the horse has not been largely replaced by petrol on large farms and light land, the writer of these lines will be much surprised. ELECTRICAL FARMING MACHINERY In France, Germany, Austria, and the United States the electric motor has been turned to agricultural uses. Where water-power is available it is peculiarly suitable for stationary work, such as threshing, chaff-cutting, root-slicing, grinding, etc. The current can be easily distributed all over a large farm and harnessed to portable motors. Even ploughing has been done with electricity: the energy being derived either from a steam-engine placed near by, or from an overhead supply passing to the plough through trolley arms similar to those used on electric trams. The great advances made recently in electrical power transmission, and in the efficiency of the electric motor, bring the day in sight when on large properties the fields will be girt about by cables and poles as permanent fixtures. All the usual agricultural operations of ploughing, drilling, and reaping will then be independent of horses, or of steam-engines panting laboriously on the headlands. In fact, the experiment has been tried with success in the United States. Whichever way we look, Giant Steam is bowing before a superior power. FOOTNOTES: 24. Cassier's Magazine. The World's Work, vol. iii. 499. 25. About HackerNoon Book Series: We bring you the most important technical, scientific, and insightful public domain books. This book is part of the public domain. Archibald Williams (2014). The Romance of Modern Mechanism. Urbana, Illinois: Project Gutenberg. Retrieved https://www.gutenberg.org/cache/epub/46094/pg46094-images.html. This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at , located at . www.gutenberg.org https://www.gutenberg.org/policy/license.html

Stories, News & Mentions about BINDERP

Binderp’s Stories on HackerNoon