Physics-based Approximation and Prediction of Speedlines in Compressor Performance Maps

Abdul-Malik Akiev; Alexander Hinterleitner; Alexander Schirger; Danyal Erg\"ur; Matthias M\"uller; Thomas Bartz-Beielstein

arxiv: 2603.11317 · v2 · submitted 2026-03-11 · 🧮 math.NA · cs.NA

Physics-based Approximation and Prediction of Speedlines in Compressor Performance Maps

Abdul-Malik Akiev , Danyal Erg\"ur , Alexander Schirger , Matthias M\"uller , Alexander Hinterleitner , Thomas Bartz-Beielstein This is my paper

Pith reviewed 2026-05-15 12:24 UTC · model grok-4.3

classification 🧮 math.NA cs.NA

keywords compressor performance mapsspeedlinessuperellipsefitting pipelineturbochargersparse dataperformance predictionoptimization

0 comments

The pith

Fitting speedlines with superellipses via a two-stage pipeline allows reconstruction of compressor performance maps from sparse measurements.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a method to approximate and predict speedlines in compressor performance maps using superellipse functions. It encodes each speedline as a compact vector of four parameters: surge, choke, curvature, and shape. A robust two-stage fitting process combines global search with local refinement to handle the data. This is validated on industrial datasets for different turbocharger types, with discussion on prediction quality for interpolation and extrapolation. The approach aims to enable full map reconstruction from limited data by incorporating physics-based elements.

Core claim

By fitting each speedline in a compressor performance map with a superellipse, the map can be reconstructed from sparse measurements and encoded as a compact, interpretable vector consisting of surge, choke, curvature, and shape parameters. The fitting uses a two-stage pipeline of global search followed by local refinement, building on prior formulations, and shows good performance on industrial data for various turbochargers.

What carries the argument

The superellipse as a functional form for speedlines, encoded by four parameters and fitted through a coupled global-local optimization pipeline.

If this is right

Speedlines can be predicted for inter- and extrapolation within the map.
Parameters provide interpretable insights into surge and choke limits as well as curve shape.
Opportunities arise for adding physics-informed constraints to improve boundary behavior.
Hybrid physics-ML mappings can be developed for full CPM reconstruction from limited data.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The method could extend to other types of turbomachinery beyond turbochargers.
Compact parameter vectors may facilitate optimization in compressor design processes.
Alternative function families could be tested if superellipses underperform on certain datasets.

Load-bearing premise

That a superellipse provides an adequate functional form for speedlines across different turbocharger types and operating conditions.

What would settle it

A dataset of measured speedlines from a new turbocharger type where the superellipse fits show systematic large errors in the surge or choke regions compared to actual measurements.

read the original abstract

Speedlines in compressor performance maps (CPMs) are critical for understanding and predicting compressor behavior under various operating conditions. We investigate a physics-based method for reconstructing compressor performance maps from sparse measurements by fitting each speedline with a superellipse and encoding it as a compact, interpretable vector (surge, choke, curvature, and shape parameters). Building on the formulation of Llamas et al., we develop a robust two-stage fitting pipeline that couples global search with local refinement. The approach is validated on industrial data-sets for different turbocharger types. We discuss prediction quality for inter- and extrapolation, metric sensitivities and outline opportunities for physics-informed constraints, alternative function families, and hybrid physics-ML mappings to improve boundary behavior and, ultimately, enable full CPM reconstruction from limited data.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper gives a workable two-stage pipeline for fitting superellipses to compressor speedlines on industrial data, but skips the quantitative error numbers that would make the accuracy claims convincing.

read the letter

This paper's core contribution is a two-stage fitting procedure that uses global search then local refinement to model each speedline in a compressor performance map as a superellipse. They reduce it to four parameters: surge, choke, curvature, and shape. This builds directly on Llamas et al. and they validate the approach on industrial datasets from different turbocharger types, with some discussion of how well it handles inter- and extrapolation. The practical side is solid. A compact, interpretable representation like this could indeed help reduce the amount of testing needed in design work, and the pipeline seems designed for reliability in fitting. The fact that they discuss metric sensitivities and possible physics-informed improvements shows they're thinking about real-world use. Where it falls short is the lack of concrete numbers. The description talks about validation and prediction quality but doesn't include specific error values, comparisons to other fitting methods, or details on how often the superellipse form breaks down for certain operating conditions or machine types. Without those, the claim that it supports good reconstruction from sparse data stays plausible but not fully demonstrated here. This kind of work is aimed at engineers in the turbomachinery field who need to work with performance maps. Someone looking for ways to approximate or predict compressor behavior from limited data would find it useful, particularly if they're already familiar with geometric fitting approaches. I would send it to peer review. The incremental advance is clear enough and the industrial angle gives it relevance, so referees could help tighten the quantitative side and check the assumptions about the functional form.

Referee Report

2 major / 2 minor

Summary. The paper claims that speedlines in compressor performance maps can be robustly approximated and encoded as compact 4-parameter superellipse vectors (surge, choke, curvature, shape) via a two-stage global-search-plus-local-refinement fitting pipeline; the resulting encoding supports inter- and extrapolation, and the method is validated on industrial datasets spanning different turbocharger types.

Significance. If the superellipse family is shown to be adequate, the approach would supply an interpretable, low-dimensional representation of speedlines that could facilitate map reconstruction from sparse measurements and downstream physics-informed or hybrid modeling; the two-stage fitting strategy itself is a practical contribution to robust parameter estimation.

major comments (2)

[Validation section] Validation section (and abstract): the manuscript states that the pipeline is validated on industrial datasets for different turbocharger types and discusses inter/extrapolation quality, yet supplies no quantitative error metrics, error bars, RMSE values, or explicit baseline comparisons; without these the central claim that the encoding reliably supports prediction remains unverified.
[Modeling section] Modeling section (building on Llamas et al.): the adequacy of the superellipse functional form for the observed range of speedline shapes across turbocharger types is treated as a provisional modeling choice rather than demonstrated; no ablation study, cross-type error analysis, or comparison against alternative families is reported, leaving the weakest assumption untested.

minor comments (2)

[Notation] Notation for the four parameters is introduced without an explicit equation reference; adding a numbered equation defining surge, choke, curvature, and shape parameters would improve clarity.
[Discussion] The abstract mentions 'metric sensitivities' but the main text does not detail which metrics were examined or how sensitivity was quantified.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments. They correctly identify areas where the manuscript would benefit from greater quantitative rigor. We address each point below and will revise the manuscript accordingly.

read point-by-point responses

Referee: [Validation section] Validation section (and abstract): the manuscript states that the pipeline is validated on industrial datasets for different turbocharger types and discusses inter/extrapolation quality, yet supplies no quantitative error metrics, error bars, RMSE values, or explicit baseline comparisons; without these the central claim that the encoding reliably supports prediction remains unverified.

Authors: We agree that the current validation discussion lacks explicit numerical metrics. Although the manuscript describes inter- and extrapolation behavior and mentions metric sensitivities, it does not report RMSE values, error bars, or direct baseline comparisons. In the revised version we will add a new subsection to the validation section containing tables of RMSE (and normalized errors) for both interpolation and extrapolation on the industrial datasets, together with error bars derived from repeated fits and comparisons against polynomial and spline baselines. These additions will directly support the claim that the four-parameter encoding enables reliable prediction. revision: yes
Referee: [Modeling section] Modeling section (building on Llamas et al.): the adequacy of the superellipse functional form for the observed range of speedline shapes across turbocharger types is treated as a provisional modeling choice rather than demonstrated; no ablation study, cross-type error analysis, or comparison against alternative families is reported, leaving the weakest assumption untested.

Authors: The superellipse form is adopted from Llamas et al. because it naturally encodes the four physically meaningful parameters (surge, choke, curvature, shape) while respecting the typical concave-down shape of compressor speedlines. We acknowledge that the manuscript presents this choice without a systematic comparison. In revision we will include an ablation study that fits the same datasets with alternative families (standard ellipses, cubic polynomials, and piecewise linear models) and reports per-type and cross-type RMSE. This will quantify the adequacy of the superellipse representation across turbocharger types and address the modeling assumption directly. revision: yes

Circularity Check

0 steps flagged

No circularity: parameters obtained by direct fit to data; predictions are genuine generalization

full rationale

The paper's core procedure is a two-stage global+local optimization that fits superellipse parameters (surge, choke, curvature, shape) directly to measured speedline points. The resulting compact vector is an output of the fit, not a re-expression of the inputs. Inter- and extrapolation claims are evaluated on held-out points or different turbocharger datasets, which are independent of the fitting data. The superellipse functional form is stated as an explicit modeling choice (building on Llamas et al.), not derived from or equivalent to the target predictions by construction. No self-citation load-bearing step, self-definitional loop, or fitted-input-renamed-as-prediction is present. The adequacy of the form is a modeling assumption whose verification lies outside the circularity analysis.

Axiom & Free-Parameter Ledger

4 free parameters · 1 axioms · 0 invented entities

The approach rests on the domain assumption that speedlines are well-approximated by superellipses; four parameters per speedline are fitted to data and therefore count as free parameters. No new physical entities are postulated.

free parameters (4)

surge parameter
Fitted boundary value for each speedline
choke parameter
Fitted boundary value for each speedline
curvature parameter
Fitted shape control for each speedline
shape parameter
Fitted shape control for each speedline

axioms (1)

domain assumption Superellipse geometry adequately represents compressor speedline shapes
Invoked as the functional form for all fitting and prediction steps

pith-pipeline@v0.9.0 · 5451 in / 1218 out tokens · 23838 ms · 2026-05-15T12:24:29.569572+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

We approximate each speedline using a superellipse... β-vector β = [ṁ_zs, Π_zs, ṁ_ch, Π_ch, CUR]^T ... polynomial interpolation to predict unknown speedlines
IndisputableMonolith/Foundation/AlphaCoordinateFixation.lean J_uniquely_calibrated_via_higher_derivative unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

The combination of DE with Nelder-Mead as solver and ortho as error metric

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.