Underhood Temperatures

[bernie at innovative.iinet.net.au]

Garfield Willis tapped away at the keyboard with:

> On Fri, 26 May 2000 09:07:35 +0800 (WST), bernie at innovative.iinet.net.au
> wrote:

> >> ..rather than get them all the way up to 800degF [no engine
> >> compartments live at those temps anyway, at least not more than once :].
> >
> >Try the exhaust manifold.

> You're not serious...well, then again maybe you are. Thanks for the
> insight, we might have missed that otherwise. But the engine compartment
> (note the words *engine compartment* above?) doesn't EVER reach anything
> near the temperature of the exhaust manifold, and nobody with even a few
> beers left in their six-pack is gonna mount electronics anywhere near
> the exh. manifold. SAE J1128 wire itself is only rated to 125degC. Ya
> know, the automotive "good stuff". Think about it.

OK -- but if you're interested in measuring the exhaust gas
temperature (or even O2) and your amplifier needs to be close
because of the weak signal, then you do need to know how hot that
gets to provide adequate thermal shielding. So the 800F (about 430C)
is an interesting temperature which you might want to measure.

That temperature becomes important when there's little convection
under the lid.

> >> This is how the big-boys do it if they can't shoot the scene with an
> >> infrared gun. If you just want a rough idea, and don't wanna do a full

> >The "big-boys" use a supercomputer to calculate the thermodynamics.
> >I won't bore you with the details.

> Good idea. I was thinking more of the race car firewall-forward
> designer, or the experimental aircraft guys designing their cooling and
> FWF installation. Those "big-boys", not the guys with white shirts and
> ties doing adverts in Detroit. These techniques are widely used in alot
> of industrial design settings as well. Computer simulations still have a
> LONG way to go (my field before I retired), especially since the most
> costly thing about them isn't so much any more the run-times or the size
> of computers required, it's the expense in building the accurate,
> detailed models, that often only a mockup or experimental setup and some
> tempilsticks or IRguns can ferret out.

I agree with the experimental verification.

But numerical analysis is becoming more accessable though and the
computing power of your Y2k desktop is about that of a 1980
supercomputer. And let's face it; that thing is sitting idle for
most of the time anyway.

Having had a great deal of experience in the automated model
generation for numerical analysis (proprietary in-house software),
I'd say that the availability of a usable product is not a technical
limitation. The methods of generating interesting nets from
dimensions are "well known".

Quite a lot has to do with politics and property rights.  No volume
manufacturer is going to give you the dimensional data of the
components you might want to analyse; not without a whole lot of
cajoling. Your only option is to measure it yourself.

> Besides, the other reason I mentioned these techniques is because of
> their accessibility to most everyone. I hazard to guess that not one

It is almost always easier and cheaper to "suck it and see" for a
specific set of circumstances.

> single person on this list is likely to use computer thermal simulation

Why shouldn't *I*. :-)

It would allow me to extrapolate to conditions which I might just
encounter; and setting up a "hot-house" is almost as involved (and
probably more expensive) then sitting down for a year of weekends
noting down dimensional and material details.

You might have access to a rolling-road place where you can crank
the thermostat up to 50C ambient on a cold winter's evening, so
different priorities might apply in your case.

-- 
Bernd Felsche - Innovative Reckoning
Perth, Western Australia
----------------------------------------------------------------------------
To unsubscribe from diy_efi, send "unsubscribe diy_efi" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org