Water Pump Housing Lightweighting: DoE-Driven Mass Reduction
The Part
The water pump housing is a pressure-containing structural casting that circulates engine coolant through a closed circuit, interfaces with the engine block via a precision bolt pattern, and supports the impeller bearing on the fluid-side face. It operates at a continuous 2.0 bar, must survive a 3.5 bar burst pressure test, and is exposed to thermal cycling from -40 degrees C at cold start to 130 degrees C at steady-state. The alloy selected for the optimized design, AlSi10Mg-T5, offers a yield strength of 240 MPa, which gave the engineering team the material basis to thin the walls and introduce directional ribbing without sacrificing the structural margin required by the pressure containment specification.
The Challenge
The legacy housing, cast in A356-T6, met every structural requirement with margin. The pressure driving redesign was not performance failure; it was unit cost in a high-volume production program where every gram removed from a casting reduces material spend, machining stock, and solidification cycle time at scale.
The difficulty was not reducing wall thickness in isolation. It was finding the specific combination of external rib geometry, wall thickness distribution, and casting process constraints that yields a lighter, castable part without triggering re-qualification of the interface geometry. A sequential manual approach, cycling through wall thickness iterations with foundry DFM review between each, can evaluate three to five concepts before program timing forces a decision. That is not enough range to find a near-optimal solution in a design space with multiple interacting parameters.
The Approach
The engineering team ran a parametric Design of Experiments across wall thickness, rib geometry, rib count, and blend radii simultaneously, generating 50+ distinct design candidates in a single parallel run. Manufacturing feasibility was not a post-optimization check; die casting constraints were applied at the generation stage, so every candidate that reached the ranking step was already castable.
The result the team did not anticipate was that the lightest castable geometry still required targeted reinforcement before it could be selected. Localized simulation identified specific zones where stress concentration exceeded the acceptable margin, and material was added back precisely there rather than uniformly across the part. The full DoE configuration, the Pareto front analysis, and the localized reinforcement logic are documented in the case study.
.avif)
Key Results
- 25% mass reduction vs. the A356-T6 legacy baseline, with all pressure containment and structural requirements met
- 4x faster engineering lead time, from approximately 3 weeks with a conventional sequential workflow to 5 days
- 50+ design variants evaluated in a single DoE run, versus 3 to 5 concepts in a conventional workflow
The case study includes the complete before/after metrics table, the Pareto front ranking across all variants, and the FEA validation data for the selected design.
Why It Matters
When manufacturability is embedded at the generation stage rather than checked after optimization, the decision the engineering team makes becomes a genuine trade-off selection rather than a filter applied to a single concept. That shift, from sequential iteration to parallel evaluation, is what makes the first commitment to a concept in a cost-down program more likely to be the right one.
Download the case study to see the complete metrics table, the Pareto variant ranking, and the FEA validation data.
Interested in reading the full case study?
Fill in the form below to instantly receive the full case study.
Aeronautics Challenges We Address
FAQs
Explore our frequently asked questions to understand how our software can benefit you.
Yes, provided the ribs are designed within casting process constraints from the start. Conformal ribs allow thinner walls to be reinforced directionally, reducing mass while maintaining stiffness and pressure integrity. The key requirement is that rib thickness, blend radii, and feature accessibility are validated against solidification and shell removal rules before the design is committed, not after.
A parallel DoE generates multiple design variants simultaneously rather than sequentially. Instead of cycling through 3 to 5 manual iterations with foundry review between each, a DoE run evaluates 50 or more candidates in a single pass, ranking them on mass, structural performance, and manufacturability. This compresses weeks of iterative review into days.
Mass reduction potential depends on starting wall thickness and the allowable stress envelope under burst pressure. In the case documented here, replacing uniform walls with a conformal-ribbed architecture in AlSi10Mg-T5 delivered a 25% mass reduction from the A356-T6 baseline, with a safety factor of 2.4 at 3.5 bar burst and all structural requirements met.
Automotive Challenges We Address
Defense Challenges We Address
Built for the Cognitive Era of Engineering
Shorten product development cycle from the earliest concept phase.
Product engineering cycle
Product engineering cycle
Reusability of engineering workflows
Unlock Your Design Potential
Request a demo to see how Cognitive Design by CDS can revolutionize your engineering workflow