Stainless, superalloys, or other??

Moderator: Mike Everman

paulengr
Posts: 6
Joined: Tue Dec 26, 2006 6:13 pm
Antipspambot question: 0

Stainless, superalloys, or other??

Post by paulengr » Tue Dec 26, 2006 10:54 pm

Before I start, I need to confess my background. I have a metallurgy degree. Second, I work at a foundry. In other words, I don't flinch watching a pulse jet get red hot. I work around stuff that exceeds 2000 F every day.

So after reading a lot of the posts, I was thinking about some ideas of ways I would try to do the same thing.

Using carbon steel is obviously nuts. It's only good to 600 F. So that immediately puts you in the stainless world.

For those who are "shocked" when they see corrosion on stainless, this is NORMAL. The temperature protection in stainless steels (ignoring some of the more exotic alloys) comes from oxidation of the chromium. Boosting chrome content is the surest and fastest way to get yourself thermal protection without going seriously exotic.

304 stainless is cheap, readily available, and limited to about 1200 F (I've seen some claims of 800 F but it can sustain 1200 F all day long for years so I don't get the 800 F claim).

316 gets you marginally to 1600 F. Structurally, it does not have the strength of 304.

The intended purpose here is that 304 is intended as "structural" stainless. 316 is for high corrosion resistance...use this for chemicals.

"Real" high temperature stainless is usually 309 or 310, with 310 being the better member. These get you all the way to about 1800 F for sustained periods. Creep is a bit of a problem, but it holds up otherwise.

To go higher, you've got to either go ceramic or to a superalloy. Inco 625 and 610 are both popular. I've worked with 625. It shapes and welds VERY easily. One of the problems with forming it is that you are limited to spin casting for shapes (melting temperature is 2400-2600 F). If you are careful about designs, you can exceed 2000 F. Over time, I've found that it has some serious creep issues.

But here's where I differ a little bit. Why not do ceramic lining? I know it sounds exotic and difficult but you can buy the materials right out of McMaster-Carr. We do it all the time and I've done several versions.

Here's my suggestion. Buy 1" hexagonal alumina (aluminum oxide) tiles that are 1/16" or 1/8" thick. These are very white and super cheap (relatively speaking). Attach them with a high temperature two part epoxy cement (again...look at McMaster-Carr). You can trivially lay them in (they come in mats to make it go fast). This gets you all the way to 2800 F with hardly any effort at all. The epoxy doesn't shrink appreciably so you can make a nice smooth surface (fill in the "grout" lines with the cement).

If you go this route, the shell can be just carbon steel or anything else. Plus as the engine heats up, the ceramic will expand faster than the steel, putting pressure against the shell and eliminating any kind of structural problems.

Above 1000 F, alumina is very strong, so forget about being brittle. Worst case is that you may have to preheat a bit to bring the engine up to temperature before reaching the point where the alumina gets some flexibility to it to withstand the shock of the engine.

Just wondering if anybody tried this yet. I know they also use tungsten carbide on spacecraft (that's what the shuttle tiles at least used to be). Tungsten carbide is great against abrasive wear but it is far more expensive than just cheap old alumina.

pezman
Posts: 613
Joined: Fri Oct 08, 2004 4:13 am
Antipspambot question: 0
Location: USA

Stainless, superalloys, or other??

Post by pezman » Wed Dec 27, 2006 12:57 am

There have been a few threads where folks have proposed ceramic pjs, and one guy built a crude one a while back.

I have had some alternative fabrication ideas but they just weren't quite viable. However, the alumina tiles might be just the thing ...

Zippiot
Posts: 1190
Joined: Tue Nov 15, 2005 7:55 pm
Antipspambot question: 0
Location: california
Contact:

Post by Zippiot » Wed Dec 27, 2006 1:03 am

What about nickel? I have hear claims that it is stronger at 1500 F than at room temp...
And its melting temp is pretty high, almost 3k right?

The icono is a nickel alloy right?
Sailing Student- How do I know if my life jacket is tight enough?
Me- Can you breathe?
Sailing Student- Yes
Me- Then its too loose!

Dave_G
Posts: 128
Joined: Tue Jun 08, 2004 4:33 pm
Antipspambot question: 0
Location: USA

Re: Stainless...Superalloys...

Post by Dave_G » Wed Dec 27, 2006 1:31 am

Are you SURE aboout Tungsten Carbide heat shield tiles covering the space shuttle?!? It's incredibly heavy (S.G. = 15.63). That's a lot of mass on a space shuttle. This doesn't seem quite right for aerospace vehicle, where every lb counts...If it's true I'll be shakin' my head in amazment...

Have you worked with 321 SS or A-286? A lot of Hiller pulsejets were fabricated from 321. 321's more costly than 304 but cheaper than the Inconels.

I like your idea of mosaic heat shield tiles, but have you experienced the vibration of a typical pulsejet? In some Hiller tests, pulsejets literally destroyed nearby insulation, adhesives and foamed-in urethanes in close proximity to operating engines, not from heat, but vibration...

The only thing they could get to stand up to the vibration were certain elastomers. Epoxy is not an elastomer.

With the right adhesive your idea is still a valid and good idea. I hope somebody here tries it and posts the results.

Zippiot
Posts: 1190
Joined: Tue Nov 15, 2005 7:55 pm
Antipspambot question: 0
Location: california
Contact:

Post by Zippiot » Wed Dec 27, 2006 1:41 am

Space shuttles has used many different things over the years, MgO bricks and Carbon Fiber I THINK are the standard...They make the MgO from dolomite and cook it at super high temps. It is light b/c mostly air is left once the CaO from dolomite breaks down.


I really wanna see a quartz pulsejet!
Sailing Student- How do I know if my life jacket is tight enough?
Me- Can you breathe?
Sailing Student- Yes
Me- Then its too loose!

Dave_G
Posts: 128
Joined: Tue Jun 08, 2004 4:33 pm
Antipspambot question: 0
Location: USA

Re: Stainless, superalloys, or other??

Post by Dave_G » Wed Dec 27, 2006 1:45 am

paulengr wrote:...Plus as the engine heats up, the ceramic will expand faster than the steel, putting pressure against the shell and eliminating any kind of structural problems.
I thought steels had a far higher coefficient of thermal expansion than ceramics, by several orders of magnitude (particularly Al2O3)? Meaning that steel will expand far FASTER than the ceramic...? Or do you mean because the ceramic is on the inside of the engine it will insulate the steel shell? And thus the ceramic will expand slightly while the steel "sees" little heat?

In your foundry experience what kind of typical temperature differentials exist between the inside and outside of crucibles and ladles? Do you use the epoxy and mosaic tile method you mention to line steel ladles with ceramic?

Mark
Posts: 10753
Joined: Sat Oct 11, 2003 10:14 pm

Post by Mark » Wed Dec 27, 2006 1:54 am

Here's some stuff I found, don't know much about tungsten carbide myself. What kind of epoxy are you referring to paulenger? There is this stuff but it's not an epoxy. I didn't think epoxy could go very high before breaking down.
http://tinyurl.com/wr4hk
http://www.afcinternational.com.au/site ... 3_tech.htm
"Tungsten carbide melts at 2870 degrees c. It (and hafnium carbide) are likely no more costly than the presently used tiles and can likely be made into a foam form simular to the present tiles. The present tiles likely resist atomic oxygen better than tungsten carbide and hafnium carbide. Hot metalic tungsten (melts 3410 c) and hot metalic hafniun burn like coal (Magnesium?) in an oxygen atmosphere, likely making both unsuitable. Neil"
"IIRC, the 'R' value of tungsten carbide is drastically lower than the current tiles. Even a 'foamed' form of that material (if possible) would be insufficiently insulating in its application, even if it did not melt."
Some tidbits.
http://www.reade.com/Particle_Briefings/spec_gra.html
321 is what they use on headers, and 347 is also mentioned as a viable pulsejet material. The inconels are good too that you mentioned.
Mark
http://www.lmengines.com/Dyno_Tuning_Seminar.htm
Presentation is Everything

paulengr
Posts: 6
Joined: Tue Dec 26, 2006 6:13 pm
Antipspambot question: 0

Post by paulengr » Wed Dec 27, 2006 5:35 pm

Zippiot wrote:What about nickel? I have hear claims that it is stronger at 1500 F than at room temp...
And its melting temp is pretty high, almost 3k right?

The icono is a nickel alloy right?
I said nickel. Yes, Inco is one manufacturer. They were the first. 610 and 625 "stainless" are almost synonymous with Inco. The trade name under Inco is called "Inconel".

Back in the 70's (I think...I forgot the exact decade), when they were trying to push the temperature limits on turbines, they turned to Inco to develop an alloy that would hold up at very high temperatures. I met one of the metallurgists that worked on the project.

Above 2000 F, the protective nature of chrome oxide pretty much gives out on you. So you are stuck with going for something else. Nickel is one way of doing it.

625 and 610 "superalloys" aren't really stainless steels anymore, but they are listed as such. The iron content is down to about 10-15%. BUT they are something like 80+% nickel.

The whole idea in reality is that it's a nickel alloy...the other elements are just there to make the nickel do something other than droop and creep.

This was the "last harrah" before they developed ways to run ports through the turbine blades where they could just do simple water cooling. Once water cooling was developed, the upper end on temperatures went from the sub 2500 F range up to the 3000+ degree range and the need for the exotic super alloys pretty much vanished.

With a pulse jet, I only see three ways of making this work. First, the fuel could be used as a coolant for the valve assembly (on valved PJ's). Second, creating a "double shell" would allow for a preheat mechanism by using the PJ's own intake air as cooling.

Third, outright water cooling could be done. Most cupolas in cupola foundries use water-cooled shells instead of refractory. You obviously lose some heat this way but it is made up in the increase in production rates due to the larger diameter working volume which overcomes the small losses through the water jacket. Oil could be an alternative (again, heat the fuel). The "diameter vs. production rate" dynamic doesn't apply to a PJ, BUT the increase in allowable operating temperature has a similar effect.

By way of example of using Inco 610, I was buying 10 foot long 4" diameter spun cast tubes made out of Inco 610 for a horizontal rotary kiln. The problem with these kilns is trying to pull a combustion gas sample BEFORE the seal in the end where it gets contaminated with atmospheric air in order to optimize the combustion process (for cement & lime kilns).

These things were very expensive ($5K-$10K as I recall), mostly subject to spot prices on the nickel market. The temperatures in that area were 1900-2100 F. In that environment, they would slowly droop and become useless every 6.5 weeks. For replacement, cut it off with a torch and let it roll out the other end of the kiln. Shove a new one in through the same port. Then we'd reclaim the spent pipe when it rolled out the other end and turn it in for a credit.

These were double-walled. There were two pipes. The inner one was just 310. The outer pipe was the superalloy. The exhaust gasses had to be cooled before sampling. So a fan pulled an air sample out of the kiln, ran it through about 20 feet of external stainless tubing (aka "air-air heat exchanger" but it wasn't anything as fancy as that sounds), pulled a small sample (at around 120 F) for the gas analyzer, and then blew the rest of the unused gasses back down around the sampling tube for cooling purposes.

I doubt anyone experimenting with PJ's is going to pay that kind of money. But these were used on kilns producing 1200 TPD of lime. Being able to measure combustion gasses meant as much as 500K BTU's per ton of production, a huge cost savings relative to the cost of the alloy.

I had considered going the refractory route for this application. It was not that hard to do. It was a matter of getting it done and the dimensional problems. All the other refractory lined equipment in that plant had a minimum of 2" of refractory on it. So I had not really experimented with seeing just how thin I could reasonably get at that time.

paulengr
Posts: 6
Joined: Tue Dec 26, 2006 6:13 pm
Antipspambot question: 0

Re: Stainless...Superalloys...

Post by paulengr » Wed Dec 27, 2006 5:41 pm

Dave_G wrote:Are you SURE aboout Tungsten Carbide heat shield tiles covering the space shuttle?!? It's incredibly heavy (S.G. = 15.63). That's a lot of mass on a space shuttle. This doesn't seem quite right for aerospace vehicle, where every lb counts...If it's true I'll be shakin' my head in amazment...

Have you worked with 321 SS or A-286? A lot of Hiller pulsejets were fabricated from 321. 321's more costly than 304 but cheaper than the Inconels.

I like your idea of mosaic heat shield tiles, but have you experienced the vibration of a typical pulsejet? In some Hiller tests, pulsejets literally destroyed nearby insulation, adhesives and foamed-in urethanes in close proximity to operating engines, not from heat, but vibration...

The only thing they could get to stand up to the vibration were certain elastomers. Epoxy is not an elastomer.

With the right adhesive your idea is still a valid and good idea. I hope somebody here tries it and posts the results.
I'm sure about the tungsten carbide. But as you said, they've rotated materials over time.

I haven't worked with 321 but I've worked a lot with 310. 310 is the stainless of choice for refractory work.

I can understand the vibration problem. There is a similar problem in designing kiln refractories. In high alumina ceramics, the problem becomes that the refractory is very brittle. Adding 2-3% P2O5 in the binder mix solves the problem. It gives the refractory a certain amount of pliability.

You simply aren't going to find anything that approximates the word "flexible" in the ceramic world. Everything is designed with expansion in mind. I didn't expect the epoxy to actually flex at all.

larry cottrill
Posts: 4140
Joined: Sun Oct 05, 2003 1:17 am
Antipspambot question: 0
Location: Mingo, Iowa USA
Contact:

Nickel Chambers

Post by larry cottrill » Wed Dec 27, 2006 5:46 pm

The combustion chambers Dr Robert Goddard's team developed in his New Mexico desert rocketry experiments were rolled and welded sheet nickel. Of course, I have no idea if he meant pure nickel in his journal or some alloy, he just referred to it as nickel. Even then, they did burn or melt through some of the chambers - usually when the distribution of liquid oxygen in the chamber went wacky for some reason. They ended up with fuel-cooled jacketed chambers, of course - one of their best innovations.

L Cottrill

paulengr
Posts: 6
Joined: Tue Dec 26, 2006 6:13 pm
Antipspambot question: 0

Re: Stainless, superalloys, or other??

Post by paulengr » Wed Dec 27, 2006 5:56 pm

Dave_G wrote:
paulengr wrote:...Plus as the engine heats up, the ceramic will expand faster than the steel, putting pressure against the shell and eliminating any kind of structural problems.
I thought steels had a far higher coefficient of thermal expansion than ceramics, by several orders of magnitude (particularly Al2O3)? Meaning that steel will expand far FASTER than the ceramic...? Or do you mean because the ceramic is on the inside of the engine it will insulate the steel shell? And thus the ceramic will expand slightly while the steel "sees" little heat?

In your foundry experience what kind of typical temperature differentials exist between the inside and outside of crucibles and ladles? Do you use the epoxy and mosaic tile method you mention to line steel ladles with ceramic?
Ladles are a totally different matter. Iron and slag are VERY aggressive in terms of wear on the ceramics. So they make everything thick and bulky. There are a few troughs though which are simply cast ductile iron with a thin ceramic coating on them. The operating temperature is usually around 300-400 degrees in the trough which is pouring 2400-2500 F liquid iron.

Yes, in bulk, expansion is a problem. But as with all things, it doesn't work that way. What you forget is that the goal is to keep the steel MUCH cooler than the hot face of the ceramic. For instance the hot faces in a cement/lime kiln will be around 2000-3000 F. The cold faces (and hence the surrounding steel) will be under 600 F, and usually closer to 400 F. This is achieved with 8" MgO brick linings. And the MgO is not used because it is more efficient than alumina or fireclay. It is simply used because the upper service temperature for aluminas is about 2500-2800, so MgO gets you into the 3500 degree range. Beyond that I have no idea what is available.

The total expansion of MgO or Al2O3 at those temperatures vs. mild steel at <600 F is such that I've had numerous problems with the opposite issue. The steel does not expand enough so that the refractory with a maximum compressive strength of about 3000-5000 PSI simply crushes. You can visibly tell when this happens because you see cracks perpendicular to the hot face and you tend to get "capping"...the faces of the bricks pop right off.

Now let's take this concern about shifting and movement a step further. In a cement kiln, the thing is so physically large that in practice treat the design like a giant water bag. It actually bulges out at the sides and in at the top due to gravity. As it rotates at the very impressive speed of 1 RPM, the shape of the wall is constantly changing.

Now here's the surprise quiz, how do they glue/cement the bricks into place?

They don't. It's all dry stacked.

In a fixed object such as a cupola or ladle, things are different. There, you can get away with cements and grouts and such. But expansion is still a serious problem. Every so many feet, you must cut expansion joints into the refractory or it will MAKE one for you as it crushes the refractory.

The end result is that on startup, most ceramic equipment starts out extremely "loose". There are air gaps everywhere. As it comes up to temperature, the gaps seal up very easily.

paulengr
Posts: 6
Joined: Tue Dec 26, 2006 6:13 pm
Antipspambot question: 0

Re: Nickel Chambers

Post by paulengr » Wed Dec 27, 2006 6:02 pm

Larry Cottrill wrote:The combustion chambers Dr Robert Goddard's team developed in his New Mexico desert rocketry experiments were rolled and welded sheet nickel. Of course, I have no idea if he meant pure nickel in his journal or some alloy, he just referred to it as nickel. Even then, they did burn or melt through some of the chambers - usually when the distribution of liquid oxygen in the chamber went wacky for some reason. They ended up with fuel-cooled jacketed chambers, of course - one of their best innovations.

L Cottrill
They make MgO (and ZrO2) components by oxygen enrichment. In a cupola, oxygen enrichment is only 2-4%. In some MgO manufacturing facilities, they use pure O2. The temperature ESTIMATES are 10,000 F. But nobody really knows since the accuracy of pyrometers at that temperature is questionable. Especially because there's no way to create a calibration standard.

The MgO itself is also the refractory...needless to say, refractory life is lousy.

Dave_G
Posts: 128
Joined: Tue Jun 08, 2004 4:33 pm
Antipspambot question: 0
Location: USA

Re: Stainless, superalloys, or other??

Post by Dave_G » Wed Dec 27, 2006 6:44 pm

paulengr wrote:The operating temperature is usually around 300-400 degrees in the trough which is pouring 2400-2500 F liquid iron.
... the hot faces in a cement/lime kiln will be around 2000-3000 F. The cold faces (and hence the surrounding steel) will be under 600 F, and usually closer to 400 F. This is achieved with 8" MgO brick linings. And the MgO is not used because it is more efficient than alumina or fireclay. It is simply used because the upper service temperature for aluminas is about 2500-2800, so MgO gets you into the 3500 degree range. ...
The total expansion of MgO or Al2O3 at those temperatures vs. mild steel at <600 F is such that I've had numerous problems with the opposite issue. The steel does not expand enough so that the refractory with a maximum compressive strength of about 3000-5000 PSI simply crushes. ... As it rotates at the very impressive speed of 1 RPM, the shape of the wall is constantly changing....on startup, most ceramic equipment starts out extremely "loose". There are air gaps everywhere. As it comes up to temperature, the gaps seal up very easily.
Very interesting and informative!

Thanks, Paulengr!

Zippiot
Posts: 1190
Joined: Tue Nov 15, 2005 7:55 pm
Antipspambot question: 0
Location: california
Contact:

Post by Zippiot » Wed Dec 27, 2006 7:24 pm

I made a little furnace that I lined with Dolomite, once I burned it out and got it up to heat there was much expansion. It doesnt expand as much anymore, but I have melted iron in it with ease!
Sailing Student- How do I know if my life jacket is tight enough?
Me- Can you breathe?
Sailing Student- Yes
Me- Then its too loose!

Mark
Posts: 10753
Joined: Sat Oct 11, 2003 10:14 pm

Post by Mark » Thu Dec 28, 2006 1:42 pm

Just a tidbit on making tungsten wire. Has anyone given thought to making a wire pulsejet? You'd have to get the woof and the warp down pat. ha
Mark
http://www.tungsten.com/tungmade.html
http://en.wikipedia.org/wiki/Woof_and_warp
Presentation is Everything

Post Reply