40 replies to this topic

[RWP80000]

Bench Racer, on 26 May 2015 - 04:42 AM, said:

We know NA engine temp is higher at the rear of the engine from history reading and the re-plumb kits available.

 

Do you by chance have an aftermarket temp gauge between thermostat and radiator? If so what are the temp diff?

 

Does anyone know the wide open throttle temp difference between the front and rear of the engine?

I have a sensor in the outlet of the radiator, and a couple of sensors in the back of the head. I've got a springfield dyno radiator, and generally see about 20-25 degF deltaT across the radiator. Spot check of a race last fall with 100degF ambient showed 200degF at the back of the head and 178degF exiting the radiator.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 

I think Jeff's data in comment #13 above is pretty reasonable.  I just completed an evaluation on a 99 NB with AIM K-type thermocouples mounted in the thermostat outlet housing and in the rear of the head in the 1/8 NPT plug in the rear of the head at the temp sensor/heater hose cover.  I ran an ICE BATH/BOILING water calibration check on both thermocouples going into the AIM logger to verfy system accuracy.  The car had a Springfield Dyno radiator, 160 degree T-stat and a heater hose looped to by-pass the heater core.  Coolant was distilled water and Redline Water Wetter at recommended mix. 

 

The bottom line is that there was about a 11 degree higher temp at the rear of the head as compared to the water coming from the outlet back to the radiator. This was true for both Max and Avg temp data throughout the laps.  Ambient temp was around 80 degrees.  Please see data summary table below.

 

On a side note I typically use a dyno ramp rate such that the length of time for the pull is close to the same time it takes our car to accelerate in 4th gear on a level straightaway.  That would be about 125 rpm per second or 20 sec pull from 4500 to 7000 rpm.  This minimizes the affect of mass acceleration and has the engine systems operating in the same dynamic rates as when on the track.

 

Date : 5-24-15   Loc : INDE  Motorsports Ranch (Az)   Car: 1999 NB #145 with 160 Deg Stat w/Sprgfld Dyno Rad using AIM K-type TC probes, Bypassed Htr. Core Driver: Steve Powers    

Lap      Time     Distance  GPS Speed   Eng RPM   Gear # A/F -avg Exh Temp #4 Coolant Temp_Rear   Coolant Temp_Outlet Fuel Temp 2 Rail Logger Tmp   Rear coolant +/- Front Outlet  

#          Min:Sec   Feet     Avg Max     Avg Max   Avg   Lambda    Max         Avg Max              Avg Max             Avg              Avg               Avg   Max

lap 2 01.54.122 11640    69.5 98.4     5419 6579 3     0.98       1171        173 180               163 172             102              94                  10   8

lap 3 01.53.682 11598    69.6 96.7     5363 7117 3     0.98       1181        179 185               168 174             108              94                  11   11

lap 4 01.56.127 11648    68.3 95.6     5228 6576 3     0.99       1189        184 195               173 186             111              94                  11   9

lap 5 01.55.651 11598    68.5 95.9     5238 6466 3     1.01       1185        189 195               179 186             114              94                  10   9

lap 6 01.53.407 11622    69.8 98.5     5375 6560 3     0.97       1192        190 195               178 183             115                          94                  12   12

lap 7 01.54.228 11641    69.5 96.9     5324 6518 3     0.98       1193        190 196               179 184             116              94                  11   12

lap 8 01.53.799 11609    69.6 97.6     5339 6479 3     0.98       1202        192 198               181 186             117              95                  11   12                                                         

                                                                                                                                    Overall Avg-> 185.3 192.0             174.4 181.6                                             10.9   10.4

Attached Files

•  99NBEngCoolantTempStudyMay2015.pdf   130.34KB   10 downloads

[ChrisA]

This one has worked well for us on many cars:

http://www.napaonlin...1666_0297059217

Mark

3x more expensive than the one from Summit though.

[Mark]

Yeah but NAPA is 1/2 mi away.

 

 

3x more expensive than the one from Summit though.

[Jeff Wasilko]

 

Bench Racer, on 26 May 2015 - 04:42 AM, said:

I have a sensor in the outlet of the radiator, and a couple of sensors in the back of the head. I've got a springfield dyno radiator, and generally see about 20-25 degF deltaT across the radiator. Spot check of a race last fall with 100degF ambient showed 200degF at the back of the head and 178degF exiting the radiator.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 

I think Jeff's data in comment #13 above is pretty reasonable.  I just completed an evaluation on a 99 NB with AIM K-type thermocouples mounted in the thermostat outlet housing and in the rear of the head in the 1/8 NPT plug in the rear of the head at the temp sensor/heater hose cover.  I ran an ICE BATH/BOILING water calibration check on both thermocouples going into the AIM logger to verfy system accuracy.  The car had a Springfield Dyno radiator, 160 degree T-stat and a heater hose looped to by-pass the heater core.  Coolant was distilled water and Redline Water Wetter at recommended mix.

 

I should add that in my case I have no tstat, which is a problem I need to correct as the car runs too cool when it's cold. The heater core is bypassed. My data above was with water/water wetter. The sensors for the data system are all AiM, though I have a AutoMeter gauge that also sends it's readings to the AiM for logging as well. I also have the ECU temp sensor in the stock location.  That also gets logged to the AiM. Belt and suspenders and duct-tape.

 

[RazerX]

I may have missed the thread but i think this was a question.  I had 1.6, with the heater core in and still plumbed.  I had two temp sensors one in the rear of the head and another exiting the thermostat on the front on the way to the radiator.  They were consistently 7 degrees apart on the track, rear sensor being 7 degrees higher than the front one.  

[Bench Racer]

Rich, appreciate the data. The Lambda/AFR ???

 

For general info. Starting after the 1.6 engine the head gasket restricted water to the frontward cylinders allowing more cooling to the rear cylinder.

[RWP80000]

Bench,

 

Lambda is used rather than the A/F ratio as the wide band O2 sensor actually measures the amount of oxygen present in the mixture for the combustion process.  A lambda of 1.00 means that there is exactly the amount of oxygen present to completely burn the fuel present.  A 0.98 lambda is indicating that there is .02 or 2% less oxygen available for compete combustion and therefore a rich mixture.  Likewise an 1.02 indicates there is 2% excess oxygen present in the mixture and is therefore a lean mixture.  The 1.00 mixture is known as the stoichiometric mixture and for typical gasoline equates to 14.67.   In reality the Bosch wide band O2 sensors are actually detecting the exhaust oxygen content and converting that to an A/F ratio based on what the gauge has been set up to represent.

 

By working with lambda you do not need to know anything about your specific fuel and you can mix your fuel with different types as the sensor is giving you the feed back on where you are relative to the oxygen content.  Maximum power is usually around 10 to 15% rich which is a lambda of 0.85.  We typically target 0.83 to 0.87 in forth gear in the 6000 to 6500 rpm range.  The average value provided in the lap by lap data set reflects all of the WOT and Closed Throttle decels but does not show any data at a specific rpm range which is what we tune with.

 

Rich Powers

[Steve Scheifler]

Might be nice if everyone talked in Lambda, but I ptretty much gave it up about a week after I got my first wideband (PLM) well over a decade ago because almost everything out there is AFR unless from hardcore tuners. But I wonder if the ideal tune (compromise between power and safe) really ends up being the same Lamda on different fuels anyway. Some have naturally better cooling, different knock resistance, etc. so if the fuels are significantly diffetent I think you probably need to know what you are tuning on anyway. For full-time closed loop operation where you aren't trying to live on the edge for every last HP (performance street car), I would definitely tune and set my target maps to Lambda, but I'm less confident that would be optimal on something like a Spec Miata. Very interested to hear about experience to the contrary though.

[RWP80000]

Steve,

I use Lambda out of convenience.  I do not have to know anything about the Stoichiometric ratio of the fuel I am running.  The stoich ratio of fuels we can legally run can vary from 14.0 to 15.0.  This means the 15% rich A/F ratio target for power would be between 11.9 and 12.75.  That a pretty good spread and if I don't have the fuel specifics I would have to test the car which is not feasible at the track unless there is a dyno there.  I use the dyno to find the fuel pressure with gives me the best power and use the lambda output as my track target.

 

The other thing is that on the coast in the cool early morning the adjusted altitude may be 100 feet while in the hot afternoon that value may rise to 2200 feet.  You would think that this would require you to lean out the car but we have found that the Miata computer overly compensates and in fact it is necessary to increase fuel pressure to maintain proper Lambda or A/F ratio.  

 

Rich Powers

[Steve Scheifler]

Yea, I know why Lamda is "better" in theory, see my very first sentence, but you still lost me in terms of testing at the track. If you can't check it at the track with a meter, then it doesn't matter which number you are working from or write down, so I must have misunderstood that part. The argument for Lambda is that when you DO check it at the track, in theory you don't need to know which fuel you have. My point, or question really, is that if I think the differences in fuel might be that big (unlikely since I run name brand street fuel) can I just assume that this very different fuel should be at the same 85% of stoich to give me max power with safety margin? Stoich is about the optimal mix to burn all the fuel, which is not our goal. I may find that my car likes 85% of that on straight gas but 86.5% on some farm-country corn blend. We are working with very blunt instruments here and as you say there are many other variables and compensation we can't (or shouldn't) control but will make both numbers drift. But I agree that in theory Lambda should keep you closer which is why I said initially, I'd be glad if everyone tuned in Lambda but they don't. If reliable locals tell me that their tuning shows that the track fuel burns pistons over 12.5 while the nearby Mobile works best at 13.5, I'll probably pay more attention to that than a % Lambda I got with what ever brew I tuned to back at home, or at least check both numbers and weigh them together to see if it tracks to my target. That was my only real point, that most people talk AFR, not which is theoretically better. That's also a convenience thing and practical reality. Mostly I want to work at minimizing the likelihood of running a fuel so different from my standard that I'll even see it.

[RWP80000]

Steve,

It's not that lambda is better than A/F, it's that to know your true A/F you need to know the stoich value of the fuel you are running.  The wide band O2 sensor is actually detecting the percent of oxygen greater or less than what is required for complete combustion.  What the gauge does to display the A/F value is to take the lambda sensor detected value and multiply it by a fixed stoichiometric constant, which for gasoline 14.67 is the typically used constant.  Therefore any displayed A/F value is going to be in error by the amount that the fuels actual stoichiometric value varies from the fixed constant value used by the gauge manufacturer.

 

I have found it is not uncommon to encounter and race with fuels with stoich A/F values ranging between 14 (race gas) and 15 (farm country fuel). So if your lambda output was 0.90 (10% rich), your gauge would typically display a 13.20 A/F but the 14 stoich 'race" fuel would actually be at a true 12.6 A/F and the 15 stoich "farm" fuel would be at a true 13.5 A/F with both being at the same 10% rich condition at these two significantly different (0.9) true A/F ratios.

 

My choice has been to eliminate dealing with the fixed arbitrary gauge constant and work with lambda targeting an acceleration enrichment constant for max power that is in the 10 to 15% range using dyno and lap time results as validation of tuning values.  I also elect to run an exhaust thermocouple in the #4 exhaust manifold runner just in case I were to encounter a gauge error or "exotic" fuel that were to show exhaust gas temps outside of normal tuning values. 

 

It is easy for many to fall into the trap of believing their gauge reading is an absolute with accuracy traceable to the National Bureau of Standards but with wide band O2 sensors it's not as easy as with your scale pads and placing calibration weights for validation.

 

Rich Powers

[Steve Scheifler]

Rich, I KNOW the difference, how it works, have forever, not my point. If you don't know your fuel, you don't really know what number you need, be it AFR or % Lambda. I'm not saying Lambda isn't closer, it is, I'm just saying you need to know the fuel regardless if you want max safe power. Which is best for the mystery fuel of the day? If our ultimate goal were to burn 100% of the fuel with no leftover O2, then of course Lambda is the whole thing. But we never want that value, and don't really care what the value is, Lambda or AFR, so long as we get the most power without undue risk. On the dyno with known fuel I determine that's 0.86. Two weeks later at the track, totally different fuel and conditions, I don't know is 0.86 is still best. Overall yes, Lambda is a better sharper tool but it's not magic because we don't want IT, we need some percent of it and THAT varies with fuel and conditions. You still end up guessing if you don't know the fuel or aren't on the dyno. And with everyone else quoting AFR you are mostly talking to yourself (as I seem to be ) so knowing AFR may be less perfect in theory but it still has some value, and for our purposes Lambda is not a panacea. All I'm trying to say.

[RWP80000]

Steve, I have no doubt that you have a good understanding of the difference in A/F and Lambda as you consistently provide good insights on a wide variety of forum discussion issues.  I am not trying to sell anyone on doing anything a specific way.  I had listed my average lap lambda and Bench questioned it.  If you believe I am trying to tune to get an average 1.00 lambda on a lap by lap basis then I have done a poor job of explaining what I am doing as I tune for a specific lambda rich of stoich at specific RPM levels in 4th gear.  I find it very easy to discuss with others what we are doing in terms of A/F, especially when running a "house" fuel as when at the Runoff's since you can bet the gauges providing the A/F are using the 14.67 conversion constant.  Personally, when I am in the position to compare and choose a fuel I will go on its heating value or RVP when running in very hot weather or high altitude situations.  And your right, we do seem to be only talking to ourselves based on the silence within this thread.

[Steve Scheifler]

:-) I'm just arguing which shade of blue the sky is anyway. I really should have said, ""Rich, I agree, I wish everyone else did, and would talk more about Lambda and less about AFR." Because that's the basic fact, but I was trying to justify why I abandoned that ideal long ago and it just kept building. I here by vow, from this point forward, to track both values when tuning on the dyno, and transition to Lambda with an occasional check of AFR onboard. The best way to spot significantly different fuel would be that ratio and it would help to eliminate some of the other variables as the cause of a change (as I think you mentioned).

Who else will join the Lambda movement?

[Bench Racer]

  And your right, we do seem to be only talking to ourselves based on the silence within this thread.

 

Not really, I'm not only listening, I am reading and I am learning. When I read your Lambda numbers of .98 -1.01 and looked at a conversion chart rather than doing the simple conversion calculation I had been for what ever reason thinking an AFR of 14.1:1 (rather than 14.7:1) was a complete burn and your Lambda numbers converted said to me, lean. Using AFR of 14.7:1 for a complete burn, not so lean. Hence my comment to you Rich, the Lambad/AFR???

 

I'm sure others are lurking and learning.

 

Rich and Steve, thank you both.
 

[Tom Scheifler]

Who else will join the Lambda movement?

Shall we print POL stickers to go with the POG stickers?

[ChrisA]

Not really, I'm not only listening, I am reading and I am learning. When I read your Lambda numbers of .98 -1.01 and looked at a conversion chart rather than doing the simple conversion calculation I had been for what ever reason thinking an AFR of 14.1:1 (rather than 14.7:1) was a complete burn and your Lambda numbers converted said to me, lean. Using AFR of 14.7:1 for a complete burn, not so lean. Hence my comment to you Rich, the Lambad/AFR???

 

I'm sure others are lurking and learning.

 

Rich and Steve, thank you both.
 

 

Same here, reading & learning. Watching AFR on-track has always concerned me that the car was running too lean. I could get the ratio on target in 4th at 6500, but with the Miata's lack of power it often takes much of a straight to get there and until it gets there she looks lean. (This also puzzles me, as one would think the car would flow too much fuel at WOT during lower RPM, then when at higher RPM, higher demand..?) How much time per lap is the engine seeing Lean vs. Rich and where/when should the "danger" flags be waving that adjustment are necessary?

 

Wish this portion of the thread could be pulled to it's own topic, It might have more interest if it wasn't flowing under Heater Core chatter. I've thrown AFR & Lamda into the topic tags for future reference.

[Bench Racer]

Wish this portion of the thread could be pulled to it's own topic, It might have more interest if it wasn't flowing under Heater Core chatter. I've thrown AFR & Lamda into the topic tags for future reference.

Chris, being you started the thread, add to the title, Tangent to AFR/Lambda Information or some such attention getter.

 

I find the AFR/Lambda interesting because of my respect for Rich and Steve and their willingness to openly share their knowledge.

 

[Ron Alan]

Who else will join the Lambda movement?

Never was a Fraternity type guy...

 

For you fuel knowledgeable guys...

 

New motor(not mine)dynoed/tuned with 91. Motor ripped...a real runner! 3rd weekend on track the owner started using 87...2nd session of the 4th weekend...loss of power on track with big rod knock when came in. Motor done! Would it be fair to say the fuel could be the soul reason for the detonation? BTW...nothing changed on the car other than fuel! 1.6 car if that matters.

[Steve Scheifler]

I think it would be fair to say that it could be main reason. Someone might argue that if tuned more conservatively it could tolerate the lower octane, but I'd ask to see their racing credentials. Of course we don't know what would have happened had he stuck with 91, it might have blown the next week. Very unwise decision either way.