← 論文一覧に戻る

高温ヒートポンプにおけるシェルアンドチューブ熱交換器のサロゲートモデル支援予備設計・実験検証・最適化

Surrogate Model Assisted Predesign, Experimental Validation and Optimization of Shell and Tube Heat Exchangers in High Temperature Heat Pumps (原題)

S. Kopparthy, Michael Lockan, P. Stathopoulos

Heat Transfer2026-06-15#省エネOrigin: EU経営インパクト: コスト削減対象セクター: manufacturing
DOI: 10.1115/gt2026-176351
原典: https://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT2026/89756/V005T13A003/7649054/v005t13a003-gt2026-176351.pdf
📄 PDF

🤖 gxceed AI 要約

日本語

産業プロセスの脱炭素化に向け、逆ブレイトンサイクル高温ヒートポンプ(CoBra)のシェルアンドチューブ熱交換器を対象に、CFDデータベースで訓練したガウス過程回帰・ニューラルネットのサロゲートモデルを予備設計ツールへ統合した研究。従来の伝熱相関式とサロゲートモデルの数値結果を実験値と比較し良好な一致を確認。さらに圧力損失を許容範囲に保ちつつ出口温度を高め、熱交換器サイズを最小化する最適化を実施した。

English

This study replaces conventional Nusselt correlations with Gaussian-process and neural-network surrogate models trained on a CFD database to predesign shell-and-tube heat exchangers in a reverse-Brayton high-temperature heat pump (CoBra) above 200°C. Surrogate and correlation results agree well with experimental data. An optimization then minimizes exchanger size while raising exit temperature and keeping pressure drop acceptable.

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

📝 gxceed 編集解説 — Why this matters

日本のGX文脈において

産業部門の熱脱炭素は日本のGX政策でも重点領域であり、電化・高温ヒートポンプの設計高度化はScope1排出削減の実装手段として参考になる。SSBJ開示とは直接関係しないが、製造業の脱炭素投資判断に資する工学的知見を提供する。

In the global GX context

Industrial heat decarbonization is a growing pillar of global transition planning, and surrogate-assisted design lowers the engineering barrier to adopting high-temperature heat pumps. It supports the electrification pathways increasingly referenced in corporate transition plans and net-zero targets.

👥 読者別の含意

🔬研究者:サロゲートモデルを伝熱相関式の代替として予備設計に組み込む手法は、相関式が存在しない構成・作動流体への拡張に有用。

🏢実務担当者:高温ヒートポンプ導入検討時の熱交換器サイズ・出口温度・圧力損失のトレードオフ設計に活用できる。

🏛政策担当者:産業熱電化を促す政策設計において、高温ヒートポンプの技術的実現性と設計最適化の進展を示す根拠となる。

📄 Abstract(原文)

High temperature heat pump systems are increasingly integrated into the process industry with the objective of decarbonizing the heat supply. To support sustainability and decarbonization goals, the DLR Institute of Low-Carbon Industrial Processes in Cottbus, Germany, has developed a high temperature heat pump pilot plant that operates on the principle of the reverse Brayton cycle (CoBra). In a heat pump system, the role of heat exchangers is critical as they facilitate the transfer of thermal energy from the source and supply heat to the industrial process. The motivation of this study is to establish a robust methodology of using surrogate models instead of analytical correlations, such that it serves as a validation to extend its application to other heat exchanger configurations and working fluids where no established correlations exist, paving the way for the widespread adoption of high temperature heat pump systems in the process industry. The current study focuses on modelling a shell and tube heat exchanger in the CoBra heat pump with sink temperature exceeding 200 °C. The heat exchanger predesign methodology adopted in this work follows a stepwise approach. First, the conventional heat transfer correlations for the Nusselt number were applied to CoBra heat exchanger data using the P-NTU approach. Second, a comprehensive CFD database was generated using 3D RANS simulations to capture detailed heat transfer behavior for different flow and geometry combinations. Third, surrogate models based on Gaussian process regression and neural networks were trained on this CFD database to replace the conventional heat transfer correlations. Finally, these surrogate models were integrated into an updated predesign tool. The numerical results from the existing conventional heat transfer correlation and surrogate models are validated against the experimental observations from the CoBra heat pump, demonstrating a satisfactory agreement. Building on this foundation, an optimization study was conducted to minimize the size of the CoBra heat exchangers while achieving higher exit temperatures and maintaining acceptable pressure drop.

🔗 Provenance — このレコードを発見したソース

🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録(無料・3キーワードまで)。

gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。