Need air @ 600-800 PSI, 1000+C ...

[MeTwo]

I need to build a source of high temperature, high pressure air for an experiment.

The air needs to be at a pressure of at least 400 PSI, but preferably 600 to 800 PSI.

The air needs to be hot as possible. 1000C+. (Sorry for mixing units...)

Right now I am envisioning a really high compression ratio single stage air compressor followed by a heat exchanger made of small diameter (<1/4" diameter) tubing.

I know that most steels have almost no strength at these temperatures.

How do I build this thing ?

Thanks

 

[metengr]

You have got to be kidding, right?

 

[racookpe1978]

How long does your experiment need to run?

That is, is it a short term rapid blast of extremely hot air that only needs to last a few seconds?

Does the supply need to be at a constant pressure, or could it be bled down from an insulated storage tank?

Or do you have a long term heat-soak type test where you need a continuous reliable supply of extremely hot air at a constant pressure for several hours/days/months?

(Note: A large volume of low pressure extremely expensive hot air is continuously available right now - but you need to be close to Washington DC to make pumping it cost effective.)

 

[zdas04]

You didn't say how much of this air you need. That matters.

I'm almost afraid to give you any of the parameters that you need to consider. Compressing atmospheric air at sea level to 800 psig is 54 compression ratios. For a water cooled machine you could do it in 2 stages. For an air cooled machine you'd want 3 stages. Doing it in one stage would give you about 630C, but no one is going to make a compressor to handle that kind of temperature and pressure rating. The practical limit is closer to 150C.

I can think of a few dozen ways that this project can kill you and everyone in the lab. Air in those conditions is just about as dangerous a substance as there is (about the only thing that is missing is that it isn't poison or radioactive). Get some help.

David

 

[MeTwo]

No, I am not kidding.

Why do I want 3 stages of compressor ? I know that intercooling the stages reduces the work that needs to be done on the air, but I need air that is hot AND pressurized, so what do I gain by intercooling ?

54 compression ratios ? I think you mean I need a pressure ratio of 54x. (800/14.7) In real life, does the ideal gas law apply to this process, the ending pressure will be the starting pressure x CR^gamma ? ie 800 = 14.7 * CR^1.4, which yields a compression ratio of about 17.5x ?

The air flow needs to be continuous. Sub 1 CFM at 800 PSI.

I am thinking of driving a single cylinder diesel engine as the compressor. Disable the fuel injector, reduce the combustion chamber volume by stuffing something in the piston crown. Put a one way valve in the glow plug hole. That should take the air from atmospheric to 600 PSI or so and make it pretty hot.

That gets me pressure, but the air still won't be hot enough at this point. So I was thinking of making a heat exchanger from 1/4" dia 310SS and putting it in an electric furnace at about 1000C. The creep rupture strength of 310SS is about 1000 PSI at 1000C. 50 thou wall thickness should do it ?

It would be great if it ran for 1000 hours, but 100 hours would work for now.

Its a continuous process. I don't need any sort of reservoir.

Thoughts/comments ?

 

[metengr]

MeTwo;
The reason for my reply above is this; you need to have an experienced pressure vessel engineer review your stated service conditions because frankly, this is not something that a typical engineer can tackle. What you are intending to build regardless of service temperature at the moment is a pressure vessel (storage or heat exchanger) and as such must be designed, and fabricated using a reputable construction code.

So, yes, are you kidding? Get some professional help.

 

[racookpe1978]

I don't recommend a thin-walled tube inside a furnace (no pressure isolation if/when anything breaks!) be the source of your hot air.

Particularly if you're going up into the red-hot (slow deforming) region for many hours: Loosing air flow (for any reason at all) means you've lost your "cooling" air through the coil that is the only thing protecting your steel from getting too hot and drooping/deforming/blowing out.

Can you heat the air in an insulated tank with an electric heatng coil or large burner as it passed thorugh the tank? If a direct-fired burner, theoutlet air from the tank would be "dirty" but it would hat up faster.

 

[MeTwo]

The volume of the "vessel" is about 200cc. It will be operating in an open field. We'll keep everyone away from it during the experiment.

310SS is OK to 1100C if the design stresses are kept down. If we run the furnace at 1100C, it should never overheat.

We can vent the furnace such that if a burst should occur within the furnace, the air rush will be vented away from the interior.

The potential energy of 200ccs of air at 1000C and 800PSI is small compared to a lot of things. Its not like we are working with a large volume of steam.

Agree/ disagree ?

 

[zdas04]

MeTwo,
Your name says you are electrical. I've been thinking about using nails and telephone wire to make a transformer to boost my houshold voltage to 7.5 kV, do I need to use real insulation or will duct tape be ok?

David

 

[vpl]

MeTwo

The potential energy of 200ccs of air at 1000C and 800PSI is small compared to a lot of things. Its not like we are working with a large volume of steam.

Disagree. Strongly. Do a search on pneumatic testing on this site. You might want to start with this one: thread378-238691 which has lots of links to pretty pictures of explosions caused by someone trying to do a pneumatic test. But keep looking, there's a lot more discussion, a lot of it real world experience that people had.

While you're not wanting to do a pneumatic test, you're trying to design a non-standard tank to supply high temperature high pressure air. If you continue, I wouldn't be surprised if pictures of your rig exploding don't show up on here one-day as yet another example of why not to play with compressed air.


Patricia Lougheed

Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of the Eng-Tips Forums.

 

[MeTwo]

We are talking about 10 feet of 1/4 diameter tubing here.

According to one of the articles, it takes 18 ft^3 of air at 500 PSI to have an energy of 2.2x10^6 ftlbs.

The volume of my "vessel" is 120 x 1/4^2 * pi/4 = 5.89 in^3.

5.89/(18 x 1728) x 2.2x10^6 = 416 ftlbs of energy at 500 PSI, neglecting delta t effects.

I'll hydro test it at room temperature.

 

[DVWE]

What is the purpose of this experiment? What are you trying to prove/solve?

 

[ApC2Kp]

MeTwo,

The source for your 800 psig air doesn't seem to be a big problem at first, if the requirement is only 1 CFM continous flow. Stainless tubing at small sizes 3/8 or 1/4 inch will handle the flow and have higher pressure ratings. The key will be to apply heating to the tubing at the very end of tubing run where it enters the reactor. There will be significant heat losses for any length between the heater and the reactor. Try to keep heated tubing length to less than 1 feet. How much power do you calculate for the air heater? The reactor is what concerns me. If the reactor is only 200 cc, then it might keep to a small diameter of less than 1/2" for 12. cu. inch. reactor 'furnace' volume?

The 1 CFM (actual? not Scfm) flow might not sound like much, but thinking about using compressed air bottles of 3,000 psig would give only a few minutes of run time at 800 psig from each air bottle. The compressor system will be substantial for 800 psig, and should also include air drier to avoid moisture problems with your experiment. The air drier would take out a fair amount of the heat of compression - don't try to save the heat of compression if you want to get the moisture out of the air. Converting the 800 psi to lbs/ft2 then it looks like 94 horsepower might be a minimum requirement for the compressor. Anybody check me on that horsepower?

 

[pmover]

MeTwo,

i DO disagree. a pressure vessel of any size containing air at the pressures and temperatures you specified is hazardous. sound advice has been given to obtain professional assistance regarding this matter and it is your choice to accept or reject.

an analogy, a fire-cracker is not something you would detonate in your hand or near human tissue without a protective barrier. trust me, the pressure vessel you so describe has more energy than that fire-cracker.

good luck!
-pmover

 

[GregLocock]

Roughly speaking you are talking about the pressures and temperatures seen in combustion gases in an IC engine. The only reason the metal parts survive there is that the heating is intermittent. So I think you being a bit blase about the material life. I'm not convinced that 316 is appropriate, SS in general is pretty disappointing.

Does the working fluid have to be air? If it had no oxygen in it (ie it was actually the exhaust from an SI engine) then you could at least look at how engines cope.

As it is very very hot very high pressure air loaded with oxygen is going to burn all the metal it touches, at a guess. Also you will find it very difficult to keep anything lubricated in that environment.



Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[bonair]

I agree with the comments about getting an engineer with experience in gas compression, heating, and safely handling high pressure gases. You need to very specifically define your requirements (flow, pressures, temperature, temperature control, duration of flow, quality of air, etc.) as a first step and "hand it over" to somebody with more knowledge. You will save time, resources, and a possible accident in the long run.

Why invent a compressor? Remember oil carry over and high temperature lead to combustion in a rigged IC engine. (as an aside, Smith Compressor of Bowling Green, KY, now out of business, made portable 150 PSIG compressors for 30+ years from a Ford V-8 gas engine with one head replaced with their special compressor head. The trailer unit was was a favorite of construction contractors for its simplicity and reliability. Point: they did not use the engine head for the compressor at 150 PSIG.) The exhaust connection is not made for 400+ PSI, do you want to be nearby when a chunk of cast iron flies by? The intermittent high pressure (and energy) in an IC is due to the explosion, not compression, that is mostly expelled as power in the down-stroke and heat.

Go the HP compressor manufacturers (Sauer, Universal Air Products, .....) and see if they have one suited for your application. Your local SCUBA tank filler has a high pressure (3 to 4000 PSIG) breathing air compressor to fill tanks. You may be able to rent a compressor from supplier or use bottled air in SCUBA or other rated portable tanks per ApC2Kp's comment Heating air should be possible by a an in-line electric heater by Chromalox or other manufacturer, rated for pressure. Since your application is fairly small-size, the heater should not be too bad. Piping and valves are a mechanical specification covered by ASME B31.3 and other standards. SS is not normally the best choice, see B31.3.

You need to first be safe. Go through a at least a couple of rethinks. Regards.

 

[hacksaw]

Awfully dangerous direction of enquiry, but good to see the cautious advice...

 

[moltenmetal]

MeTwo: knowledgeable people have done things far more dangerous than you're planning and survived it. That doesn't mean you'll be in that company though.

If you persist, it's better that you do so with advice than without. Consider this advice rather than ecouragement-

If all you're after is a short-term experiment, bottled gas is the way to go. Forget about procuring a compressor for a short-term experiment. And even if you go with a compressor, forget about getting any benefit from the work of compression at this tiny flow. Heat loss will eat it all.

As to your question, you need the intercooling because the compressor materials won't survive if they're run too hot. This question on your part has me wondering if you know enough about this stuff to survive this experiment.

You won't successfully operate a 1000 C electric furnace in "an open field", so you're being a bit disingenuous- you at very least need a roof over this thing.

You'll also need to have whatever is receiving this hot gas inside the same furnace. At these conditions, a few inches of tubing will lose you 100 C regardless how well you insulate. For your preheater, consider putting an ARI Firerod heat trace inside your tubing, with the air flowing in the annulus between the heat trace and the tubing ID. They won't last long heating air to 1000 C though!

http://www.ariindustries.com/heaters/heaters.php3

The 309 or 310 SS tubing is nice in theory, but the trouble will be the hot end connection. Forget about Swagelok-type compression fitting connections- they're not available in 310SS, 316SS won't cut it for long, and that fitting design is not going to hold together at these conditions to the rupture strength of the tubing anyway.

Figure out what your hydrotest pressure would be at the ratio of allowable stresses between room temperature and 1000 C and you'll get an idea of what you're up against.

Bottom line: you need cone and thread connections or the like, or a very risky weld. Have a look at Autoclave Engineers.

You've given no indication of what's on the downstream end of this. No valve is going to survive throttling a small flow of extremely hot air- you'll need to cool first.

 

[MeTwo]

"Figure out what your hydrotest pressure would be at the ratio of allowable stresses between room temperature and 1000 C and you'll get an idea of what you're up against."

Very good idea. Thanks.

 

[moltenmetal]

Be careful when actually DOING that though- codes limit the actual test pressure to avoid yielding the components when they're cold...