Underhood Temperatures

[Garfield Willis]

On Fri, 26 May 2000 10:15:57 +0800 (WST), bernie at innovative.iinet.net.au
wrote:

>You change the system in order to observe how it performs. 
>All you'll onserve is how the changed system performs. Or
>don't the temp sticks have any physical properties at all?
>
>If I crack open a chip and paint the surface with the pen then try
>to observe the temperature - I'll probably get very misleading
>results in relation to a practical application.

Non sequitur, dood. If you paint a swipe of T-stick across the chip
surface and then heat it up, you can DEFINITELY tell if it's reached the
T-stick temp or not. Your example of cracking open a chip is way off
point. Something's cracked alright, but it's not the chip.

>It's the same situation as placing a strain-gauge on a mechanical
>component - the strain-gauge carries part of the load. All you can
>do is assume that the strain-gauge isn't significant. Your
>measurements should tell you that so that you can estimate your
>error i.e. the degree of uncertainty.
>
>The Heisenberg principle applies to all systems which are observed.
>It's an early awareness of chaos.

The chaos... I can see that. None of the above amounts to anything of
consequence in the real world we've been discussing. The T-sticks are
like crayons; you're tellin us we ought to worry about the heat effects
of a swipe of crayon on some object inside the engine compartment. Get
real.

>A couple of seconds to a minute more likely; on my engine a couple
>of minutes makes a world of difference to the inlet manifold (which
>is directly above the exhaust manifold).

OK, in your case don't mount any electronics near your intake manifold,
cuz it's near your exhaust manifold. Simple, huh?

>The relative temperature changes are most significant and depend
>hugely on engine layout and the operation of ancillaries. (Electric
>cooling fan.)
>
>You might get more time with a mechanical cooling fan.

When you shoot underhood temps with a gun, you don't have EITHER fan
running. You shut down, let soak or no, then open the hood and start
probing. Other than thin metal baffles and the like, everything else
VERY slowly coasts down. Changes in just a few seconds you just don't
see; sorry, it just doesn't square with practice.

>You're assuming that the thermal capacity of the piece of aluminium
>is the same as for the components. It might be applicable for
>steady-state application, but then again, your piece of aluminium
>isn't generating any heat either. All you'll get is a baseline
>ambient temperature before your own thermal contribution.

Nope, not assuming anything of the kind; just tryna get close enough for
that cigar. The purpose of the sample isn't to predict the eventual temp
of the electronics when they DO produce heat, it's just to get a decent
take on the ambient temps in that spot. Once you do that, you can then
use the standard engineering analysis process I mentioned earlier in
this thread to predict temp rise and see if that location will give you
the margins you need to be confident it will work out.

>Yeah... pretty neat gadgets. I saw one which had a laser pointer but
>the sales guy couldn't verify that the laser spot was even pointing
>at the spot being measured. I didn't even ask about the power output
>of the laser. :-)

Good idea, you'd have had to wonder about how much heat the laser
pointer put into the thing your measuring then, huh? Ever try it on your
hand? Heh. Besides, I'd never look to a sales guy for my technical
understanding anyway. The typical way to align the beam is to use a very
hot point source, like the end of a hot metal rod, and scan the area to
peak; the beam should be pointing at the hot point source. Quite
practical and repeatable.

>> Oh, to answer your question, *I've* heard of Heisenberg. Wasn't he the
>> guy that built that blimp that crashed? So, what about him? :)

>Perhaps what he was saying was lost in the translation.

Well, lost definitely. Perhaps not in the translation tho. Thanks for
the chat; I almost feel like I've been to some cocktail party where they
strain to discuss the most inconsequetial things they can possibly think
of. You'll excuse me, I think I'll go look for the hors d'oeuvres.

Gar


----------------------------------------------------------------------------
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