dlaing Posted April 10, 2009 Posted April 10, 2009 Fascinating, alright -- but wrong. I've been testing various metals in the conducting stud of the low mass GM sensor, including brass and copper so far, and just got a ROUND TUIT on this, as I was preparing to thread an aluminum stud for testing. Checking the numbers before I started making aluminum shavings, it turns out there's no need. Thanks again for bringing the concept of thermal diffusivity to my attention, Dave. However, as it turns out, it isn't a significant consideration here, because, contrary to your incorrect post above, the numbers follow the same trend from metal to metal as thermal conductivity (see below). As a rule, assuming CREDIBILITY is your objective -- may I (again) recommend that you NEVER, EVER, EVER trust Wikipedia as an orignial source -- in this case, there's no online link to the original source to be found there, but instead, a listing of a reference book with data evidently not accessible online. This doesn't make the original source INVALID, it just means it isn't accessible online for purposes of verification. As another rule, assuming CREDIBILITY is your objective, which is NOT ALWAYS THE CASE (certainly hereabouts ), or necessarily anywhere else for that matter -- may I also recommend that you ALWAYS ALWAYS and ALWAYS pay attention to both UNITS OF MEASURE, and ORDERS OF MAGNITUDE, both of which you've ignored above, which has resulted in the above ERROR of perception on your part. If we take a hypothetical and assume the above Wiki numbers are valid: Copper has higher thermal diffusivity than aluminum. From your Wikipedia number above for copper: 1.1234 × 10^-4 = 11.234 × 10^-5 So WRT Thermal Diffusivity in m²/s (again, please always use correct units of measure), when properly matching up ORDERS OF MAGNITUDE: Copper 11.234 × 10^-5 m²/s Aluminium 8.418 × 10^-5 m²/s By all means, if you believe it is Moi who've got this wrong, please challenge this with your best understanding (and backup -- not citing Wikipedia as an original source). I've found it very hard to find thermal diffusivity numbers of all 4 metals under my current consideration online for comparison purposes. If you find an orignial, verifiable source -- please advise. FWIW, here are the RELATED, BUT NOT EQUIVALENT thermal conductivity numbers I'm using, from an original source, verifiable online, which matches up with numbers on other original sources, also verifiable online: Thermal conductivity copper -- 231 BTU/(hr-ft-°F) aluminum -- 136 BTU/(hr-ft-°F) brass -- 69 BTU/(hr-ft-°F) mild steel – 26.0 – 37.5 BTU/(hr-ft-°F) SOURCE: http://www.engineersedge.com/properties_of_metals.htm Using Wiki's number for a common brick: Common brick 5.2 × 10-7 is equal to 520 x 10-5 which is greater than either the copper or aluminum And for air: Air (1 atm, 300 K) 2.2160 × 10-5 It is actually the lower number that we want. So, I guess air has a better number than copper or aluminum. So, it looks like the highest OR lowest thermal diffusivity is NOT exactly what we want. Not Wiki's fault but mine. Diffusivity being the ratio of thermal conductivity to volumetric heat capacity is not exactly what we would want. Air would be ideal if it was dense enough to raise the conductivity number, which won't happen on this planet. Aluminum still is a better metal than copper or brass, but that is assuming the same volume. Certainly one could make a smaller part out of copper that would work better, due to copper's superior strength. High conductivity and low mass is the bottom line, which air has the best ratio, but that alone does not make it the best material. What would work better than aluminum or copper is aluminum with a copper or silver core to do increase conductivity, but that may be overly complex. Again, why not put the sensor directly in the head? I still have not heard a convincing answer. Anyway, through your experimentation I think you realize that the ultimate conductivity is not necessarily what we want, otherwise you would not be running that GM sensor with an air gap, right? Keeping the mass down is what makes the GM sensor work well. Aluminum is lighter than Copper and that is what gives it an edge, IMHO. (and wiki's numbers back it up)
Guest ratchethack Posted April 10, 2009 Posted April 10, 2009 You can lead a horse to water. . . . . . [sigh]. . .
Skeeve Posted April 10, 2009 Posted April 10, 2009 Drrr... Ratchet is correct, thermal conductivity is the essence of the matter here, as far as getting the heat that the sensor is supposed to sense to that said sensor... Dave is correct in that there may be other concerns, such as weight or mass, involved in the selection. WRT Dave's question why the sensor can just be mounted directly: don't forget, the more embedded in the head the original sensor was, the more likely it was to be sensing it's own heat via conduction thru the sensor body, rather than the tip. Given how kludgy the entire factory-bodged inclusion of this sensor was [as has been determined during the course of the prior 40-something pages in this thread], it's not too surprising that we still require some sort of stand-off for the GM sensory bodge, given its thermoplastic housing. FWIW, the ideal material from a performance/weight standpoint would be beryllium, but since it has other distinct handicaps such as expen$e and toxicity, I'll leave any disputes over my assessment to back-channel chatter... Pure [well, pure enough to be called copper rather than bronze or brass] would be the ideal material for conducting the heat from head to sensor, given the limited nature of our application. Let's all just agree on that & move along, O.K? Ride on!
Guest ratchethack Posted April 10, 2009 Posted April 10, 2009 Given how kludgy the entire factory-bodged inclusion of this sensor was [as has been determined during the course of the prior 40-something pages in this thread], it's not too surprising that we still require some sort of stand-off for the GM sensory bodge, given its thermoplastic housing. Aye. I think of the necessity of an air gap (as it serves the function of smoothing out the response of the thermistor as presented previously here multiple times by Dan M and Yours Truly) as a buffer -- a simple enough concept, but can you imagine attempting to sell this to all those who can't follow the most thuddingly simple and basic thermodynamics and heat flow concepts, as demonstrated hereabouts of late. . .?? How about selling the wisdom of using an air gap as a buffer, coupled with the use of the most highly conductive base possible for the sensor holder?? That should go over big around here, too. How heavily conflicted d'you reckon the most vocal few hereabouts would get over that one?? The mind reels. . . Why, it's all gibberish, don't ya know? Pure [well, pure enough to be called copper rather than bronze or brass] would be the ideal material for conducting the heat from head to sensor, given the limited nature of our application. Let's all just agree on that & move along, O.K? OK by me. Analysis of exactly this^, with a low-mass GM sensor, backed up by real-world testing on the road to follow (some time after the rain stops).
Greg Field Posted April 11, 2009 Posted April 11, 2009 Don't forget to anal-ize the options for the perfect buffer gas to fill the gap. There are untold gains to be found there, and surely another 47 pages . . .
dlaing Posted April 11, 2009 Posted April 11, 2009 Pure [well, pure enough to be called copper rather than bronze or brass] would be the ideal material for conducting the heat from head to sensor, given the limited nature of our application. Let's all just agree on that & move along, O.K? Ride on! Sorry, my vote is still for Aluminum to replace the brass adapter, or the body of the sensor, assuming the same volume. But sure, drill a hole and tap in a copper or silver bolt to approach the air gap This horse can be led to Kool-Aid® but he ain't drinkin'. The Sheople can fall for that. Anybody weigh the brass adapter? How does it compare to the mass of the sensor? I will agree that in the plastic adapter copper is the better choice since there is not much mass to effect thermal inertia. But surely the brass adapter would be much better off made of aluminum than copper. The conductivity is only needed for the short path from cylinder head to probe, assuming you add conductive, as if we even need that much conductivity with stock mapping!!!! But yes, for those running rich enough to benefit on a warmed up engine from adding thermal goo, you might actually benefit from replacing your brass adapter with a copper one and your plastic and copper alloy OEM adaper with a core of PURE COPPER!!! Checkout the lousy conductivity numbers for "copper brass" which is probably what he OEM plastic adapter utilizes. http://www.engineeringtoolbox.com/thermal-...tals-d_858.html Of course someone could probably take the brass adapter to a lathe and shave off half it's weight, but it still won't conduct like aluminum. Weather resistance is about all the brass has going for it.
dlaing Posted April 11, 2009 Posted April 11, 2009 How about selling the wisdom of using an air gap as a buffer, coupled with the use of the most highly conductive base possible for the sensor holder?? That should go over big around here, too. Yah, if we are talking a copper or aluminum replacement for the brass housing, choosing copper over aluminum to increase the conductivity that minimal distance and then using an air gap as a buffer does baffle the mind. Other scenarios it is trickier to determine if copper or aluminum is ideal. If you are talking the threaded rod from cylinder head to the possibly heat damaged piece of plastic, housing the GM sensor, I don't think a pure aluminum rod with less or no air gap will behave very different than pure copper rod with more air gap. It would depend on the air gap. I don't think you could easily calculate what gap with copper would have the same conductivity as aluminum with no gap, since in either scenario you have to heat the air around the GM sensor's thermistor, not just the thermistor or the air in the gap. If you ignored the air and housing that would be heated, the air gap would be less than a hundredth of a millimeter, but that is strictly theoretical with very crude math. In the case of the OEM plastic housing, the thermal inertia of copper or aluminum would be pretty similar, by itself, but with a heavy sensor like the OEM, I agree the conductivity of copper trumps the lower mass aluminum, and copper is the better material.
Dan M Posted April 11, 2009 Posted April 11, 2009 In the case of the OEM plastic housing, the thermal inertia of copper or aluminum would be pretty similar, by itself, but with a heavy sensor like the OEM, I agree the conductivity of copper trumps the lower mass aluminum, and copper is the better material. Really, I think Ratchet has hit on the best temp sensor set up for this non feed back system on an air cooled engine. An air temp sensor actually sensing air temp (yes, I know the air is not moving). Beyond that, the difference between aluminum and copper and brass for a stud is just splitting hairs. The system is not that particular. Getting past the enormous lag of the factory set up and having an air gap at the same time pretty much covers it. Not that anybody ever splits hairs here...
Skeeve Posted April 11, 2009 Posted April 11, 2009 But sure, drill a hole and tap in a copper or silver bolt to approach the air gap Actually, this has some merit, if the thread is fine enough: you can adjust the air-gap to achieve the general characteristic of heat flow to the sensor tip as you desire... As for the best gas to use, that's easy: radon, as it is the heaviest noble gas available! Of course, there's that whole radioactivity downside... Hmm, shall we stick w/ xenon? Damn the $$$, full -liness ahead!
Guest ratchethack Posted April 11, 2009 Posted April 11, 2009 Actually, this has some merit, if the thread is fine enough: you can adjust the air-gap to achieve the general characteristic of heat flow to the sensor tip as you desire... FWIW, the way I've been adjusting air gaps for testing has been twofold: 1. progressively shortening up the "business end" (the end toward the thermistor) of the conducting studs, while maintaining the OE measurement of 10 mm for the external part that threads into the head. 2. running all, part, or none of a "short stack" of o-rings under the collar of the GM/NAPA Echlin sensor, which is a push-in, interference fit in the Delrin holder I'm using. So far, with either brass or copper stud, there hasn't been any detectable difference on the road between any air gap tested yet >3 mm between the stud and the protective cage around the thermistor. With gaps under 1 mm or so, the dastardly low RPM sensor body heat soak hot lean-burn feedback loop symptoms begin (just begin, mind you) to raise their ugly heads after hauling down from fully warm, hard riding at extended stops. Yep, as Dan has noted, I'm afraid this bodes poorly for future expectations of thermo-goop in this setup. WRT that, I defer further comment until the infamous ROUND TUIT makes a future appearance.
al_roethlisberger Posted April 11, 2009 Posted April 11, 2009 HOLY COW.... 47 pages for a head temp sensor I actually would like to know the ideal reliable street setup for this once my plastic holder finally bites the dust one day, but can we summarize? Drop me an IM when it's settled Al
dlaing Posted April 11, 2009 Posted April 11, 2009 Really, I think Ratchet has hit on the best temp sensor set up for this non feed back system on an air cooled engine. An air temp sensor actually sensing air temp (yes, I know the air is not moving). Beyond that, the difference between aluminum and copper and brass for a stud is just splitting hairs. The system is not that particular. Getting past the enormous lag of the factory set up and having an air gap at the same time pretty much covers it. Not that anybody ever splits hairs here... I agree with regard to his setup! Except that Brass piece he is using could surely be improved upon by either PURE Aluminum or PURE Copper. (splitting hairs I would choose aluminum, if I could find it easily because it has less mass to heat up and more than adequate conductivity) After that, he just has to find the ideal air gap and make sure the plastic assembly holds up to his infamous Mojave runs.
dlaing Posted April 11, 2009 Posted April 11, 2009 Actually, this has some merit, if the thread is fine enough: you can adjust the air-gap to achieve the general characteristic of heat flow to the sensor tip as you desire... As for the best gas to use, that's easy: radon, as it is the heaviest noble gas available! Of course, there's that whole radioactivity downside... Hmm, shall we stick w/ xenon? Damn the $$$, full -liness ahead! I think my laughter in that quote was for the next line down...my bad.
dlaing Posted April 11, 2009 Posted April 11, 2009 HOLY COW.... 47 pages for a head temp sensor I actually would like to know the ideal reliable street setup for this once my plastic holder finally bites the dust one day, but can we summarize? Drop me an IM when it's settled Al I can't wait till we get to the air temperature and pressure sensors!!!
dlaing Posted April 11, 2009 Posted April 11, 2009 Really, I think Ratchet has hit on the best temp sensor set up for this non feed back system on an air cooled engine. An air temp sensor actually sensing air temp (yes, I know the air is not moving). More Hair Splitting: EUREKA!!!!! The Guzzi air temperature sensor has IDENTICAL output to the "oil" sensor temperature. So it should be a better solution for the those that aren't content with possibly 10% deviation. But, I don't think it would fit the existing wire harness, or does it? And the GM generic part is cheaper. The GM sensor fits the existing wire harness, right? Remember when Skeeve discovered the factory reversed the seat and "choke" springs? Wouldn't it be funny if reversing the sensors worked better! How many of us would be bothered by delays in measuring air temperature with a heavy brass coolant sensor? Probably not all the people with K&N pods and dangling air temperature sensors. That worth reading through 47 pages for!!!!
Recommended Posts