Firewall seperate from seat?

[MikeWaggoner at UW]

Correction:
"The Tg is the temperature where a material starts to flow. It is a common term for polymers. For example polyethylene, ABS, PVC, and polypropylene also have a Tg. The matrix material is commonly a polymer (epoxy or polyester) and these generally have a low Tg."

Tg is NOT the melt temperature James. PE, for instance, is utilized above it's Tg (which is well below room temp). Tg is the temperature at which the plastic switches between brittle ("glassy") and rubbery behavior.

(ps - message edited to make Denny look crazy)

[This message was edited by MikeWaggoner at UW on March 11, 2004 at 07:33 PM.]

[Denny Trimble]

Do you mean PE is utilized above its Tg?

[James Waltman]

Yeah sorry Mike, I know. Wrong choice of words. Thanks for clarifying it for me.

[Denny Trimble]

I think all the dihydrogen monoxide in this corner of the states is getting to us...

[dancin stu]

Having used wet lay phenolic resins before with glassfibre, I would not recommend it to anyone! It is horrible stuff to work with. Unlike polyester, it does not break down the binder between the fibres, and so is reluctnat to conform to any shape, also the fumes are vile.

Produces nice, bright orange components though!

[Jarrod]

I know its not specifically on the same topic, but I was surprised at the Oz comp this year at the number of teams that had no firewall between the fuel tank and exhaust. I know the rules don't say anything about it, but for peace of mind surely you would run one? Any one have comments?

[Michael Jones]

Learn something new everyday...this time it's useful for a change .

On that note, I've always found Tg applied to the weave and resins in tandem confusing - so, if you have a low Tg resin on fiberglass, is the resin the weakest link in a heat situation? If so, using standard epoxies without heat involved in postcure still will cause havoc, no?

Depends on what is done with respect to the seat and firewall combination, but I'm envisioning from the initial post something that will need to take considerable force (e.g., a driver strapped in and pressed back into said seat/firewall at 1g or so.) Something reaching rubbery consistency back there sounds like it will deform with this load, rendering the seat/firewall useless regardless of it being Kevlar or 'glass, if indeed the resin is the weakest link.

Lecture away, I'm a bit rusty learning about this but it's always interesting stuff.

[Garbo]

We did a quick burn test comparing Wet lay epoxy-CF with a 1/2" aluminum honeycomb core to 0.032" aluminum sheet. We used a Propane Torch for heat, 2" away from the surface.

Disclaimer: This test was not very scientific and was as much a late night "Let's play with fire." as anything.

Structurally, the aluminum was affected very little after 5 minutes. The aluminum did heat up to several hundred degrees (6 inches from center of flame, on 'driver side' of the sheet) in seconds.

The epoxy burned out of the hot side of the sandwich, starting 10 seconds or so after the flame was applied. The result was bare, brittle carbon cloth on the hot side. The temperature on the cool side never changed more than a couple degrees in 5 minutes and the cool side laminate which appeared to have unchanged properties.

The thought at the time was "I'd rather be sitting in the carbon seat." on the other hand, the rapidly heating aluminum seat might prompt you to eject, even if the track workers were asleep.

Garbo

[James Waltman]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Garbo:
The epoxy burned out of the hot side of the sandwich, starting 10 seconds or so after the flame was applied. <HR></BLOCKQUOTE>

Garbo,
Seems like a good reason not to use it as a firewall. You have all of the correct elements there just not in the right combination. Use aluminum skins on your aluminum honeycomb and you'll have the best of both worlds.

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Michael Jones:
Lecture away, I'm a bit rusty learning about this stuff but it's always interesting stuff. <HR></BLOCKQUOTE>

Michael, you asked for it.

The resin is almost always the weakest link for a fiberglass composite part when heat is involved. S glass has a max melting point of 1725C. I think that it gets soft well before that but you get the idea. A resin that is good to 100C on a fiber that is good to 1000C. As an example of the high temp capability of fiberglass: we are using some E glass weave as temporary header wrap.

Standard epoxies without a post cure would be hard pressed to stand up to engine temps. As one of our professors is fond of saying "it is all about the crosslinking"�. Epoxies are thermoset polymers that cure by crosslinking (condensation reaction). During the cure process the viscosity of the resin increases once the molecule(s) start to crosslink. When the viscosity increases it makes flow more difficult and it makes it more difficult to continue crosslinking. When you post cure (or cure) a part at an elevated temp you give the resin some extra energy (heat). This allows some movement and allows more crosslinks to form. Once the crosslinks have formed they stay. The higher the crosslink density the higher the Tg and generally better mechanical properties. The resin's ultimate Tg is when the maximum number of crosslinks have formed.

Polyester resins are a little different. They cure by an addition reaction. This is like a chain reaction. That is why it only takes a little bit of initiator to cure large quantities of polyester. Polyester resin contains a monomer. The purpose of the monomer is to form crosslinks. Typically the monomer is styrene. Styrene gives polyester the smell so many of us have grown to love. Chlorostyrene is used as a monomer in some resins to help improve the flame retardant properties. At room temperature the polyester gets about as crosslinked as it ever will.

Thermal cycling can be a real problem in composites. Raise the temp, lower the temp and repeat. Sounds a lot like fatigue. The temperature swings can set up large cycles of internal stress. This can lead to problems even without load on the part. It is probably not a real concern for any of our applications but I'm on a role...

TGDDM is a good epoxy for high temp applications but is a nightmare to work with (think hot melt glue). It has been used here for one of the projects that required a resin system that could stand up to high temps. It is the part in the middle of the following picture. This picture is a little old and a bit more progress has been made. I put this picture up on our website a long time ago and hid it in with pictures of the V8 (that's the V8 in the back). I figured it would generate some interest but nobody asked me about it until Tim Heinemann did a few weeks ago.



This summer I got hired on as a research assistant to help build the body for Viking 32. Viking 32 is a full size car that we are building for the Dept of Transportation. The body is all carbon fiber. We post cured the whole thing once we were done. I will see if I can post some pictures later of our set up for post curing Viking 32's body.

Hope I didn't slip up and make any mistakes in this rambling. Feel free to call me on it if you find something.

[Michael Jones]

Once again, I learn something new every day, and once again, it's interesting. Thanks much! It at least makes logical sense based on what I have discovered in my scanning the literature and filled in a few gaping holes in my knowledge of composite mechanics.

We're not particularly strong in that here generally speaking, and I'm as much of an engineer as the doctors on ER are doctors, formally speaking, so I'm bound to miss a few details.

On thermal cycling, what temperature would this start to matter? I would assume at the lowest Tg of the combination of resin/curing technique and fibers used.

Given that you don't see this as much of concern, I suppose the number of cycles to failure is too high for us to care much, given the special-purpose nature of these cars. That noted, you do suggest the E-glass header wrap is temporary, which suggests that number is not, say, the fatigue life cycle of steel.

I would also guess that cycles until failure will decrease if it did bear a load of some magnitude.

So, in our firewall/seat example posed, would this come into play? There's heat (and localized heat - I'd imagine that's an issue as well, since parts of the glass wall would be much cooler than others, near headers for example) and by definition the application of the heat is cyclical, but there's not all that many cycles over the course of use in a year. And there's a force pushing back...

Hm. I'm beginning to be glad that we didn't do this three years ago. Even if it worked, we would've had no idea why.