The Vortex-Killing Elbow Joint
Why we built it
This project began with a practical engineering problem.
Electromagnetic flow meters perform best when the incoming flow is stable and fully developed. Unfortunately, real piping systems rarely provide ideal conditions. Elbows, valves, pumps and other fittings introduce swirl, secondary circulation and velocity distortions that reduce measurement accuracy and often require long straight pipe runs before the meter.
We wanted to know whether a constraint-first design methodology could discover a better solution.
The result is the vortex-killing elbow joint.
The design challenge
Designing an effective flow conditioner is an extremely difficult optimization problem.
Every small geometric change influences multiple competing objectives:
Optimizing one objective often makes another worse.
Traditional optimization methods therefore spend much of their effort searching through designs that never had a realistic chance of succeeding.
A different approach
Our design engine approaches the problem differently.
Rather than searching directly through millions of possible geometries, it first builds an internal structural representation of the problem.
That representation identifies:
- Structural constraints
- Admissible design families
- Equivalent geometries
- Dependent parameters
- Invariant relationships
By removing impossible or redundant candidates before optimization begins, the engine searches a dramatically smaller and more meaningful design space.
The exact implementation is proprietary.
The principle is straightforward:
Reduce the search space before attempting to optimize it.
Why the geometry matters
The internal geometry does not attempt to predict fluid flow directly.
Instead, it helps organize the engineering problem.
Several features proved especially important.
Constraint propagation
A single design decision often affects many apparently unrelated properties.
The geometry exposes those relationships early, allowing impossible designs to be discarded before simulation.
Parameter collapse
Many variables that appear independent become linked once the structural constraints are applied.
Instead of optimizing every geometric feature independently, the engine optimizes a much smaller number of meaningful structural decisions.
This dramatically reduces the effective search space.
Symmetry reduction
Many candidate designs are functionally identical despite looking different.
The geometry groups equivalent solutions together, allowing the optimizer to evaluate one representative rather than rediscovering the same solution repeatedly.
Topological organization
The engine distinguishes between superficial geometric changes and changes that fundamentally alter the flow-conditioning mechanism.
This allows optimization to proceed within structurally useful design families instead of wandering into unrelated regions of the search space.
Scale separation
Some properties depend strongly on operating conditions.
Others are structural properties of the design itself.
Treating these separately improves both optimization efficiency and the transferability of promising designs.
Conventional engineering remains the final judge
The internal geometry is never used as a substitute for engineering simulation.
Every candidate design is compiled into a conventional CAD model and evaluated using established computational fluid dynamics.
Throughout development we relied on widely used open-source engineering software, including OpenFOAM, for high-fidelity CFD analysis. The design engine used simulation feedback to calibrate its search, but every significant result was validated using conventional CFD rather than internal scoring alone. (VortexFlowMeters.com)
In other words:
The geometry proposes designs.
Conventional engineering evaluates them.
Current status
The vortex-killing elbow joint is currently undergoing physical bench testing.
The experimental program is intended to verify the computational predictions under controlled laboratory conditions.
The encouraging result so far is that the measured behavior is tracking closely with the simulations, increasing our confidence that the improvements are real rather than artifacts of the optimization process.
Additional testing across operating conditions and installation configurations is still underway.
Why this matters
The elbow joint is important for more than its own performance.
It demonstrates the original purpose of this research.
The internal geometry was developed as a better way to organize complex engineering problems.
The flow conditioner is one example where that methodology produced a commercially useful design.
The quantum-computing work described elsewhere on this site was never intended to become a product. It was developed as an independent stress test of the same design methodology. Seeing similar improvements emerge in both fluid dynamics and quantum engineering gave us confidence that the underlying building blocks were doing something genuinely useful.
That unexpected convergence is why this work continues.
Every improvement to the building blocks improves the design engine.
Every improvement to the design engine creates opportunities to discover better engineering solutions.