- Rating
- 2
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
This paper proposes a pretraining method for fluid dynamics prediction using unlabeled 3D geometries. The method employs the Oseen equations as proxy dynamics and uses a physics-informed loss for pretraining, without requiring precomputed source simulation fields. Experiments show that the pretrained model improves prediction accuracy across multiple fluid dynamics benchmarks.
Critical strengths
- The paper presents a coherent geometry-based pretraining framework that incorporates simplified fluid physics and improves downstream prediction performance.
- The ablation study supports the benefit of the coupled constitutive and momentum constraints: the full objective outperforms a variant retaining only incompressibility and pressure-trace consistency.
Critical weaknesses
My main concerns lie in the experimental evaluation.
- The improvements over the strongest competing methods, typically GeoPT, are relatively modest, and the test sets of the four smaller benchmarks contain only 10–44 cases. Without results across multiple random seeds or uncertainty estimates, it is difficult to determine whether these gains are robust. Reporting repeated runs, particularly for the proposed method and GeoPT, would make the comparison more convincing.
- The experimental configurations of the baselines are insufficiently documented, making it difficult to assess the fairness of the comparisons. In particular, the GeoPT model variant, pretraining and fine-tuning settings, and whether its results were reproduced under the same evaluation protocol should be clarified. The introduction emphasizes GeoPT’s lack of explicit fluid coupling, whereas Figure 1(b) emphasizes its offline data-generation cost. These are distinct motivations: the paper should clarify their respective contributions and provide a comparable cost breakdown to substantiate the claimed efficiency advantage.
Questions for authors
-
The paper motivates the Oseen equations as a simplified proxy retaining key fluid interactions. What makes this choice particularly suitable for transferable pretraining compared with alternatives, such as Stokes or the full Navier–Stokes equations? Is there analytical or empirical evidence supporting this trade-off between physical expressiveness and training difficulty?
-
What justifies characterizing DrivAerNet++ as “data-rich” beyond its larger dataset size? To support the claim that the benefits extend beyond label scarcity, could the authors provide learning curves comparing training from scratch and pretrained initialization across different training-set sizes on this benchmark?
9 comments
Q1: The paper motivates the Oseen equations as a simplified proxy retaining key fluid interactions. What makes this choice particularly suitable for transferable pretraining compared with alternatives, such as Stokes or the full Navier–Stokes equations? Is there analytical or empirical evidence supporting this trade-off between physical expressiveness and training difficulty?
Our goal is not to solve high-fidelity Navier–Stokes equations during source pretraining, but to construct condition-dependent proxy tasks that retain essential physical interactions with a tractable residual structure. This motivates our choice of Oseen.
First, unlike Stokes, Oseen preserves the explicit dependence of the volumetric velocity equation on Reynolds number and background-flow direction. With , the equations become
Under our velocity boundary conditions and zero-pressure reference, rescaling pressure removes the explicit Reynolds-number dependence from the continuous Stokes velocity equation. By contrast, Oseen retains the interplay of background advection and viscous diffusion. This makes Oseen better suited to generating condition-dependent supervision for the same geometry across different flow regimes.
Second, relative to full Navier–Stokes, Oseen retains essential transport and coupling terms while eliminating the nonlinear disturbance self-advection term. The momentum residuals satisfy
Oseen retains background advection, pressure coupling, and viscous diffusion, thereby preserving condition-dependent transport structure. Full Navier–Stokes adds the disturbance self-advection term . With fixed geometry, physical conditions, and discrete operators, the Oseen residual is linear in the predicted fields and yields a quadratic least-squares objective, whereas the nonlinear self-advection term introduces higher-order field couplings into the Navier–Stokes loss. This structural simplification motivates Oseen as a proxy for large-scale pretraining without claiming a guarantee of faster neural-network optimization.
The additional Stokes experiments provide empirical support:
| Dataset | Scratch | Stokes | Oseen (Ours) |
|---|---|---|---|
| Aircraft | 0.0922 | 0.0587 | 0.0405 |
| DTCHull | 0.1876 | 0.1520 | 0.1362 |
| NASA-CRM | 0.1091 | 0.0938 | 0.0866 |
| Car-Crash | 0.1971 | 0.1833 | 0.1762 |
Stokes pretraining improves upon Scratch on all four tasks, confirming that even simplified proxy physics can provide useful supervision. Oseen reduces the error further by approximately 31.01%, 10.39%, 7.68%, and 3.87%, respectively, supporting the value of background transport for downstream transfer.
Together, the theoretical comparison and Stokes experiments support Oseen as a principled compromise between condition-dependent physical expressiveness and residual complexity. The Stokes comparison tests the value of background transport empirically; the Navier–Stokes comparison establishes the structural effect of omitting nonlinear disturbance advection.
Q2: What justifies characterizing DrivAerNet++ as “data-rich” beyond its larger dataset size? To support the claim that the benefits extend beyond label scarcity, could the authors provide learning curves comparing training from scratch and pretrained initialization across different training-set sizes on this benchmark?
DrivAerNet++ provides a substantially richer supervised setting than our smaller benchmarks through both geometric diversity and dense spatial field annotations. Its 5,817 labeled training geometries span different vehicle body types, wheel configurations, and underbody designs, with pressure and wall shear stress (WSS) fields evaluated at 32,768 surface points per geometry. We therefore use “data-rich” to describe the scale and diversity of the available high-fidelity supervision, not to imply that performance has saturated.
With the complete labeled training set, our method achieves relative L2 errors of 17.91% for pressure and 18.51% for WSS. The corresponding literature-reported scratch-trained Transolver errors are 23.87% and 22.49%, giving relative error reductions of 24.97% and 17.70%, respectively. These results demonstrate competitive performance on a substantially larger benchmark and broaden the empirical evidence beyond small-label tasks.
To characterize how pretraining benefits change with increasing target supervision, we are conducting additional learning-curve experiments across different label budgets on DrivAerNet++. Given the computational cost of repeated training on this benchmark, these experiments remain in progress, and we will report the results once available.
Critical Weakness 2(a): The experimental configurations of the baselines are insufficiently documented, making it difficult to assess the fairness of the comparisons. In particular, the GeoPT model variant, pretraining and fine-tuning settings, and whether its results were reproduced under the same evaluation protocol should be clarified.
We use GeoPT's officially released eight-layer pretrained checkpoint (GeoPT_8layers.pt), with a hidden dimension of 256, eight attention heads, and 32 slices. During downstream adaptation, we reinitialize the task-specific output layers while retaining the remaining pretrained parameters. Rather than copying results from the original paper, we independently fine-tuned and evaluated GeoPT on all four downstream benchmarks.
To ensure a fair comparison, GeoPT and our method use identical dataset splits, labeled subsets, input preprocessing, prediction targets, and evaluation metrics. Both are fine-tuned for 200 epochs with the same optimizer and checkpoint selection protocol. Relative L2 errors are evaluated per test sample after restoring predictions to the original scale. Thus, the comparisons are conducted under a unified downstream protocol.
Critical Weakness 2(b): The introduction emphasizes GeoPT’s lack of explicit fluid coupling, whereas Figure 1(b) emphasizes its offline data-generation cost. These are distinct motivations: the paper should clarify their respective contributions and provide a comparable cost breakdown to substantiate the claimed efficiency advantage.
We agree that these motivations should be distinguished. Our primary methodological contribution is learning geometry–condition–field relationships through explicit proxy-physics supervision, whereas GeoPT learns geometry–motion representations from synthetic geometric dynamics. By directly enforcing governing-equation residuals during pretraining, our approach additionally avoids offline synthetic trajectory generation.
To quantify the computational benefits, we compare source-stage investments and downstream costs under a common accounting framework. Our measured preprocessing time is 7 h 15 min 52 s, and pretraining requires approximately 127.48 GPU-hours, including data-loading overhead. GeoPT reports approximately 80 CPU cores for 3 days to generate around 5 TB of synthetic trajectories, followed by 144 GPU-hours for Base pretraining.
Using illustrative rates of USD 0.02 per CPU-core-hour and USD 2 per GPU-hour, we assume 200 CPU-core-hours for our preprocessing. For a downstream task at matched accuracy, we assume 100 labels and 6 GPU-hours for Scratch, versus 60 labels and 3 GPU-hours for both pretrained methods, with each high-fidelity label costing USD 2.56.
| Cost component (USD) | Scratch | GeoPT | Ours |
|---|---|---|---|
| Source preprocessing and supervision generation | 0 | 115.20 | 4.00 |
| Source pretraining | 0 | 288.00 | 256.00 |
| One-time investment | 0 | 403.20 | 260.00 |
| Downstream geometry processing per task | 5.00 | 5.00 | 5.00 |
| High-fidelity labels per task | 256.00 | 153.60 | 153.60 |
| Downstream training per task | 12.00 | 6.00 | 6.00 |
| Total per downstream task | 273.00 | 164.60 | 164.60 |
| Break-even task count | — | 4 | 3 |
Under these assumptions, our method requires 35.5% less upfront investment than GeoPT and reduces recurring downstream costs by 39.7% relative to Scratch. Consequently, our pretraining investment is recovered after 3 downstream tasks, compared with 4 for GeoPT.
Although these monetary estimates rely on illustrative assumptions, they demonstrate the potential benefits of eliminating offline supervision generation and amortizing physics-aware pretraining across downstream tasks.
We thank the reviewer for the detailed concerns about experimental reliability and the choice of proxy equation. We have added multi-seed and Stokes comparison experiments and clarified the methodological rationale, computational costs, and scope of the large-scale benchmark results. Our responses follow.
To assess training variability, we performed three independent downstream fine-tuning runs for each of our method and GeoPT on all four fully labeled tasks, holding the pretrained checkpoints fixed and varying the fine-tuning seeds. The table reports mean ± standard deviation of relative L2 error.
Our method yields lower mean errors on all four tasks under the fixed benchmark splits, with the largest improvement on Car-Crash. These repeated runs quantify fine-tuning variability, not uncertainty arising from the limited number of test geometries; our conclusions are therefore confined to the evaluated splits.