Business, Space

PDR or CDR?

21 Comments

I’m glad that NASA has approved SpaceX’s Preliminary Design Review for the launch abort system, but this doesn’t seem right:

NASA has approved the preliminary design review of SpaceX’s launch abort system, which will help crews aboard the Dragon capsule get out of harm’s way should any problems crop up during liftoff, company officials announced last week. With this hurdle cleared, SpaceX can now start building hardware for the system.

Emphasis mine. That’s not my understanding of a PDR. I thought that PDR meant that you could start doing detailed design, advancing to a Critical Design Review. After passing CDR, you start bending metal. Is there not going to be a CDR?

21 thoughts on “PDR or CDR?”

  1. I read recently that SpaceX plans to conduct the first test of their launch escape system in December. No other details were provided but I suspect they’re talking about a LES rocket static test. If they’re really that close to a test, it sounds more like a CDR than a PDR unless it’s just a proof of concept verification.

      1. I read that as a test of the “Super-Draco” engine to be used on the LAS, not a full test of the LAS itself.

  3. The issue isn’t NASA approval so much as whether there has been a successful CDR. If you’re building hardware before you’ve done a CDR, you’re defeating the purpose of a CDR.

    1. I think I see the point, but it should be all right to build subassemblies or test hardware that retires the largest risks. PDR is “if this engine works as we expect (and we _think_ it should, because this and that), then…”. And CDR is the same thing, only “we know that it works because we tested it”.

      Russian system has even 1 more stage: after their equivalent of CDR, there is a milestone of “Documentaton Produced”. This Documentation is delivered to the manufacturing plant, which is supposed to be able to gear up and start stamping rockets and spacecraft like sousages. Anything that is made at the Shop that is attached to Design Bureau does not count for the real hardware, even if it flies in orbit.

  4. There are a lot of programs where if you know what some of the COTS items are going to be, or if you have in house hardware, it is perfectly reasonable to cut metal in order to reduce risk and increase confidence in your eventual CDR. For example, for our work, we would go ahead and order the Hall thruster even before PDR or the solar cells or even build test panels as we know what we want and what is is going to take to get there. Why waste time and schedule by not ordering or building the hardware you know that you are going to need.

  5. A CDR gives the go-ahead for production hardware to be built, at least in Air Force parlance. Pre-PDR hardware is built all the time. I even saw one contractor assert he had qualified his propulsion system before his proposal was submitted (unfortunately he won that competition, over my vociferous protests. Then began a long journey of discovery about what “qualifiied” actually means).

  6. Both PDRs and CDRs are exercises in evaluation. One can certainly evaluate based only on analysis and simulation from first principles. But it’s easy to imagine that just bending some metal and trying it out is more convincingly dispositive, faster and cheaper than a pure analysis and simulation approach in many cases. Also, as noted elsewhere here, it gives one a leg up on the eventual building of production hardware. This reported approval is, indeed, excellent news.

  7. I agree with MfK. If you haven’t built hardware by the CDR, you’re in trouble. Heck, if you haven’t prototyped the main technical risk components by the PDR, you’re in trouble.

  9. My thoughts on how this went:

    SpaceX (optimistically or not): We’re going to test our escape system next year.

    NASA: What!? We haven’t had a design review yet! A CDR is required before you start building.

    SpaceX: Well, we’ve already started building, otherwise we wouldn’t be able to test next year.

    Another part of NASA: When was the PDR?

    First part of NASA (after lots of emails and directory searches): There wasn’t one.

    Another part of NASA: Have to have a PDR first!

    SpaceX: Fine. Let us know what to call it, and we’ll show you what we’ve done. We’ll bring our cool videos.

  10. I’ll throw my hat in with many others. PDR says you accept the design concept. CDR means you are ready for production. Between PDR and CDR, you need to test the design.

    To me, the success of SpaceX is that they are not afraid to test prior to CDR or even PDR. What they appear to do is fine safe places to test, where failures will not cause loss of life and property. I say this as a person involved with NASA abort test. After the safety assessments were complete and the range secure, another year and several million dollars went by as the design was perfected. The point of the test was to discover flaws without the excessive expense of analysis. Despite the requirements for the test, it was treated as a design validation test. This increased cost and schedule, while providing little since one test well within envelope hardly validates anything.

    Here’s to SpaceX moving forward and getting a test done in the near future. With the holiday’s approaching, I’ll consider them a roaring success if they launch by February.

  11. I think this conversation nicely illustrates primary reasons for the cost and development time differences between NewSpace companies and legacy aerospace 🙂

  12. Tom Hill: speaking of cool SpaceX videos, looks like SpaceVidcast.com is finally back online. I wonder when they are going to announce that SpaceX bought SpaceVidcast?  I heard that this happened a couple of months ago. I am curious as to whether there be exclusive SpaceX stuff on this site and if the service be free or if SpaceX be charging for it.  I hope it is free. Does anyone know?

  14. This is what Spacevidcast people were saying. Were you in the Tweetup tent? I was. You act like this is your blog.

Comments are closed.