Modelling for the LDSD parachute - what are the issues with scale models and software simulations?

Question

The LDSD¹ parachute is too large to test in any current wind tunnel, the only test possible on the ground was the rocket sled test.

In this NASA Edge episode (starting at 15:12), Ian Clark talked about modelling the SIAD², saying it is a rigid body and so scale models can be tested.

Why aren't scale models of the parachute useful for testing?

Did software modelling help at all with the design? At 10:15 of the press conference for LDSD this week, Clark talked about how limited the data for supersonic parachutes was before the first LDSD test, and how much the data from it improved understandings of the dynamics. Are we at all close to being able to create a software simulation for that?

Because really, how do you design a supersonic parachute without such tools?³

---

^{1 Low Density Supersonic Decelerator
2 Supersonic Inflatable Aerodynamic Decelerator
3 That question is perhaps not really answerable briefly. But I just had to ask.}

Answer 1

The scale model of a vehicle with a SIAD for use, for example, in a ballistic range is made out of steel. This is ok, because the SIAD will not significantly change shape in flight. The shape of a parachute on the other hand changes a great deal in flight, so a rigid model will not represent the behavior even close to correct. Any attempt to make a small flexible model will fail, due to our inability to scale down fabric proportionally. The fabric used in the full scale parachute is already not far from as thin as you can make it.

As for software simulations, no, we are nowhere near being close to making a realistic simulation of the reaction of a parachute to a supersonic flow and simultaneously the reaction of the supersonic flow to the parachute.

Software modeling is used in a limited way, assuming for example that the parachute doesn't change shape and using the simulation to determine the loads on the fabric and seams for that particular shape. Then we can do that with a different shape. Then we apply factors on that for the actual materials and seams used, since we know that approach will underestimate the real loads.

Scale models were also used to compare different configurations for drag and stability, in our case at about 1/3 scale (so it would fit in that big wind tunnel). However, we knew that we could only get comparisons that way, not absolute estimates for the full-scale parachute.

Because really, how do you design a supersonic parachute without such tools?

We do what we can, apply robust margins, and in the end it comes down to trial and error by test. It is not possible today to design a high-performance, supersonic parachute on paper, and have any confidence that it will work the first time. That is the unfortunate reality.