So That is a Regen Engine on SpaceX's Bench
Michael Belfiore had some pictures that Elon allowed him to take during a trip to SpaceX's factory during the ISDC meeting last week. One of them in particular caught my interest:
The caption said that it was the "Merlin 1C, the nextgen SpaceX engine coming together on the bench." When I looked at the picture, I thought to myself "My, that looks like copper, I wonder if it's a regen version of the Merlin." Then I figured that naw, it was probably just the lighting angle. After all, if you look at the the inside of their ablative nozzles, they do look a bit copper colored:
But I figured there was no harm in asking. Well, according to Michael, my guess was correct, this was a regen version of the Merlin that SpaceX is working on. Now, from the picture it wasn't clear if it's a tube-wall type design, or if this is a milled-channel wall, or if it was something different entirely. It'll be interesting to see how it turns out though. With how little trouble we had with regen cooling, and since their engine is substantially larger than ours, only slightly higher chamber pressure (about 40-50% higher), and uses a propellant that tends to leave a nice insulative carbon soot layer on the inside of the chamber, I'd be surprised if they had much trouble at all with the cooling. In fact, they'll likely be able to get several seconds of Isp back by being able to retune their injector for optimal performance. And the engine will be a lot more reusable. Of course, their turbopump will need to raise the kerosene pressure a bit higher, but with what the optimal O/F ratio is for kerosene, it probably won't increase the pump propellant consumption very much at all. All in all a smart move for SpaceX.
That alone should greatly increase the odds that a Merlin engine failure on a Falcon 9 is sufficiently "graceful" so as to allow for real engine-out capability. The one failure they had last year that damaged the test stand (the one that got leaked all over the internet) probably wouldn't have been possible with a regen engine. There's a chance that there still could have been a burnthrough in some extreme situation, but those tend to be something you can detect before they get out of hand and shut the engine down in time, unlike the massive structural failure that is possible with an ablative type nozzle.
Just speculatin', but it sounds like a really good move on their part.
The caption said that it was the "Merlin 1C, the nextgen SpaceX engine coming together on the bench." When I looked at the picture, I thought to myself "My, that looks like copper, I wonder if it's a regen version of the Merlin." Then I figured that naw, it was probably just the lighting angle. After all, if you look at the the inside of their ablative nozzles, they do look a bit copper colored:
But I figured there was no harm in asking. Well, according to Michael, my guess was correct, this was a regen version of the Merlin that SpaceX is working on. Now, from the picture it wasn't clear if it's a tube-wall type design, or if this is a milled-channel wall, or if it was something different entirely. It'll be interesting to see how it turns out though. With how little trouble we had with regen cooling, and since their engine is substantially larger than ours, only slightly higher chamber pressure (about 40-50% higher), and uses a propellant that tends to leave a nice insulative carbon soot layer on the inside of the chamber, I'd be surprised if they had much trouble at all with the cooling. In fact, they'll likely be able to get several seconds of Isp back by being able to retune their injector for optimal performance. And the engine will be a lot more reusable. Of course, their turbopump will need to raise the kerosene pressure a bit higher, but with what the optimal O/F ratio is for kerosene, it probably won't increase the pump propellant consumption very much at all. All in all a smart move for SpaceX.
That alone should greatly increase the odds that a Merlin engine failure on a Falcon 9 is sufficiently "graceful" so as to allow for real engine-out capability. The one failure they had last year that damaged the test stand (the one that got leaked all over the internet) probably wouldn't have been possible with a regen engine. There's a chance that there still could have been a burnthrough in some extreme situation, but those tend to be something you can detect before they get out of hand and shut the engine down in time, unlike the massive structural failure that is possible with an ablative type nozzle.
Just speculatin', but it sounds like a really good move on their part.
posted by Jon Goff at 10:37 PM
11 Comments:
I'd really like to know if Flometrics pump solutions would scale up to Merlin with comparable performance to their current turbopump. That would _really_ decrease the odds of inflight self-disassembly.
Kert,
I'd really like to know if Flometrics pump solutions would scale up to Merlin with comparable performance to their current turbopump. That would _really_ decrease the odds of inflight self-disassembly.
Steve seems to think he could actually do better than Merlin's turbopump both for total dry mass and for Isp hit from gas generator fuel consumption. Of course, he hasn't ever flown one of his pumps, but from what he was saying at Space Access, that may change sooner rather than later.
~Jon
Juan,
Does the design target of matching the performance of a turbopump fed rocket using their design require pushing rocket pressures up to the 500 psi range, or will they get same/better performance at any chamber pressure on the basis of overall system mass and the ability to run the tanks dry?
I'm not entirely sure, but 500psi is still not that big of a hassle. We haven't had any real problems with it. My understanding is that gas generator turbopumps, since they have to burn some of the propellants to generate the pump power, end up cutting the "effective Isp" of the engine, and the turbopump itself, plus the extra tank mass for the turbopump propellants can be rather heavy.
I think though that in order to get better than GG turbopump performance was assuming supercritical helium with a heat exchanger. If you use high pressure gas, the system ends up being lower performance than a turbopump...but even with the supercritical helium/heat-exchanger system, you're still talking about a much, much simpler and cheaper system than any realistic biprop turbopump.
Possibly completely reusable liquid rocket first stage. Man, I'm looking forward to this.
Btw, why doesn't Ambivalent Engineer allow comments? He posts all kinds of cool engineering stuff I'd love to talk about.
Jon, I got a few more details on the Merlin 1C from Tom Mueller, SpaceX propulsion chief. He says it will have 88k lbs. of thrust at sea level, same as the 1B. Also, it's of milled copper construction.
Why copper?
Thermal conductivity?
Mike
Thermoconductivity is probably a big part of it. It also tends to distribute hot spots a lot easier so you're less likely to get a burn-through. We use copper for many of the same reasons. In fact, it's probably (I could be wrong) the most common material for using in making regen chambers. It's reasonably strong in some alloys, keeps its strength to a much higher temp than aluminum, has really, really good thermoconductivity, and there's a lot of experience with it.
If you have a system that has a high enough heat flux to not want to use aluminum, and you don't use copper, what else would you use? I guess you can use stainless or one of the superalloys, but those ain't cheap.
~Jon
If you have a system that has a high enough heat flux to not want to use aluminum, and you don't use copper, what else would you use?
Not to be practical, but zirconium has a higher thermal conductivity than aluminum, and a much higher melting point. Tungsten's thermal conductivity approaches that of aluminum, and has a very much higher melting point.
Paul,
Not to be practical, but zirconium has a higher thermal conductivity than aluminum, and a much higher melting point.
You sure about Zirconium? Looking on Matweb, they say it has a thermal conductivity of only ~16.7W/m*K, which is much lower than aluminum.
Tungsten's thermal conductivity approaches that of aluminum, and has a very much higher melting point.
Now, on this one you're right. I'll be darned. Way too dense for practical rocket purposes, but it is quite strong, and probably loses almost no strength over the likely temperature range. Not to mention that fabrication would be a flaming pain in the backside. But interesting, you learn something new every day. I had always assumed that Tungsten had a low thermoconductivity.
BTW, what do you think about my post about the molten floride thorium reactors?
~Jon
Hmm. I was looking at this chart, which shows Zr having a high thermal conductivity. But looking at my CRC handbook, I guess the person putting together that chart dropped a decimal point.
About the Flometrics pump. The supercritical helium system was flown on the lunar lander. Ariane 5 uses liquid helium to pressurize its LOX tank.
I talked to Tom Mueller and Elon about the pump in 2002, but Tom had already built a gas generator by then and the Flometrics pump was not running at the time, it was only a concept. It is a lot cheaper than a Turbopump, but it still takes time and money to design one.
Steve Harrington
Post a Comment
<< Home