Charge Temp Approximation Table, FD RX7

[Pete_89t2]

I'm using the modeled fueling method on my G4+ Fury, and am looking for guidance on how to properly set up this table. I understand the math behind how this table is used to compute a charge temp from the IAT and CLT inputs, but was interested if anyone had some empirical data for the FD to share to inform these table inputs.

In my case, I have a fast reacting GM IAT sensor installed in the throttle body elbow, and I also log data from a 2nd Bosch air temp sensor that is installed in the OEM location in the upper intake manifold (UIM). Generally, these 2 sensors track and agree with one another to within a couple of degrees F, but after heat soak condition (e.g., shut the car off hot & restart ~ 10 minutes later), the sensor in the UIM will initially read about 10~20*F higher than the IAT sensor in the TB elbow, and after a few minutes of driving both sensors temp readings will drop, and they will match temps again within a few *F.

Given the above observations, I'm thinking my charge temp approximation table should be populated with zeros for all RPMs and MAP values at idle or above, and perhaps no more than 10~15% at MAP values below idle for all RPMs. My rationale here is since the intake path from the TB elbow to the engine's intake port is relatively short, the air charge won't receive much heating beyond its measured IAT value at the TB or UIM.

Thanks!

[Adamw]

I dont have one for a Rotary but I would expect still to be some heating from the hot manifold and cooling jacketed intake port at low air speeds.  Less effect than a piston engine where the intake charge is passing over a valve that is a few hundred degrees.  

You can start at zero, get a small area of your fuel map tuned so that lambda is on target, then increase air temp (while coolant temp is relatively constant) and see if mixture stays on target.  Sometimes you can do this by blocking the intercooler with cardboard or using something like brake ducting on the turbo to suck hot air off the exhaust manifold or similar.

[Pete_89t2]

19 hours ago, Adamw said:

I dont have one for a Rotary but I would expect still to be some heating from the hot manifold and cooling jacketed intake port at low air speeds.  Less effect than a piston engine where the intake charge is passing over a valve that is a few hundred degrees.  

You can start at zero, get a small area of your fuel map tuned so that lambda is on target, then increase air temp (while coolant temp is relatively constant) and see if mixture stays on target.  Sometimes you can do this by blocking the intercooler with cardboard or using something like brake ducting on the turbo to suck hot air off the exhaust manifold or similar.

 

Thanks Adam, I think I have a strategy for tuning this table now. Worst case for heat transfer from the engine block to the intake air would be when the air is stagnant or not flowing rapidly thru the manifold, so I think idling or cranking to start after a heat soak would represent a worst case condition. Here's the test I came up with, based on your advice above - please comment if I'm missing something:

1. Set the entire charge temp approximation table to zeros

2. Get car up to full operating temperature and let it idle while logging my CLT, IAT, UIM air temp and lambda (measured & target), with closed loop lambda turned off. Use this data to baseline how well measured lambda tracks with lambda target at idle speed.

3. Use a heat gun to blow hot air over the intercooler core - this should raise the IAT & UIM air temp values fairly quickly.

4. While doing #3, watch the lambda's to see how much the measured lambda deviates from target.

5. If there's minimal to no deviation, I can leave the charge temp table set at all zeros, otherwise start incrementing the charge table cells around idle by a small amount, say 5% each step, while watching lambda's to try to null out the deviation to close to zero.

[Adamw]

I highly doubt a heat gun will do anything.  You will need a much larger heat source.

[Jefferson.C]

I'm only familiar with RB engines but I have found starting with the Thunder base map example charge temp table has always worked well for me. 

[Jefferson.C]

As for IAT heat soaking I have found using a 4D table like this has helped cure lots of my personal RB engines for restarts with heat soaked IAT and fuel rail, where the engine would run very lean momentary until the heat soaking would cool off. I didn't like just increasing the values too much in the post start enrichment table at operating temps because I only get heat soaking after 10+ mins of being turned off. No need for rich restarts if I just have the car off for a short period of time. The post start decay time table at operating temps is a little longer, around 10 seconds. It worked for my applications but I'm sure it could have been finessed but gives you some ideas.

 

 

[hugemikeyd]

hey @Pete_89t2, i know this is an old topic by now, but i was just curious if you were able to get some better numbers? I feel like on my RX-7 the exhaust and general heat held within the intake manifold are influencing the actual temperature of the air. I noticed that if my Charge Temp gets too warm, the car has trouble idling and will stall quite often, so I am looking for other ways to set this table properly.

I have watched this Webinar: https://www.hpacademy.com/previous-webinars/075-charge-temperature-compensation-link-g4/

Andre seems to take advantage of a dyno, which i don't have easy access to, so i was hoping to do this in my garage or with some logs from a drive

[Pete_89t2]

On 6/14/2023 at 9:08 PM, hugemikeyd said:

hey @Pete_89t2, i know this is an old topic by now, but i was just curious if you were able to get some better numbers? I feel like on my RX-7 the exhaust and general heat held within the intake manifold are influencing the actual temperature of the air. I noticed that if my Charge Temp gets too warm, the car has trouble idling and will stall quite often, so I am looking for other ways to set this table properly.

I have watched this Webinar: https://www.hpacademy.com/previous-webinars/075-charge-temperature-compensation-link-g4/

Andre seems to take advantage of a dyno, which i don't have easy access to, so i was hoping to do this in my garage or with some logs from a drive

 

I did manage to get a charge temp table sorted out, which has worked well for all conditions except for a heat soaked start up & idle. In that case, I'll still experience some idle instability for a minute or two after a heat soaked startup. To correct that, I've been thinking of adding a 4D fuel trim to tweak my fueling only under heat soaked start up conditions, so the 4D correction table axes would be something like charge temp vs time after engine start, and corrections would null out to zero after say 1 minute of run time

I also watched that HPA webinar you linked, and I'm also lacking a dyno. So I did it on the road by analyzing my log data to iteratively improve the charge temp approximation table. Since I had a 2nd IAT sensor fitted in the upper intake manifold (UIM), I used data from logged that sensor, as well as logs from my faster responding IAT sensor (located in the charge pipe), and logged my coolant temps, RPMs & MGP/MAP data to do some data comparison under all conditions to tweak the table. Goal was to get the charge temp that the ECU calculates as close as possible to the 2nd UIM IAT sensor's reading under steady state driving conditions (i.e., when you would not expect to see much change in IATs).

[hugemikeyd]

would you mind sharing what your table looks like? trying to get a sense for a proper shape

[Pete_89t2]

3 hours ago, hugemikeyd said:

would you mind sharing what your table looks like? trying to get a sense for a proper shape

Looks like my attachment limit is blown, but here's a quick cut & paste of the charge temp approximation table from my tune file. Top row is at MGP = -15psi, from 0 RPMs (left most column) to 7000 RPMs (right most column). Next row down is at MGP = -7psi, 3rd row down is at MGP = 0psi (transition point to positive boost). All the rows under that are in positive boost (MGP = 7psi, 15psi & 22psi). FWIW, my FD is running a stock ported 13B-REW with a 13B-RE Cosmo upper & lower intake manifold, GM 90mm DBW throttle. Port matching was done to match the REW ports to the Cosmo LIM runners. It has an efficient large core IC in the stock mount configuration with dedicated cold air ducting, and I fabricated a cold air intake to supply fresh air to the turbo (single BW SXE 300 series). My logged IATs typically run no more than 20~25*F higher than whatever the ambient air temp is.

8    7.5    7      6.5     6.5    6      5.5    5
6    5.5    5      4.5    4.5    4      3.5    3
3    2.5    2.5   2       2       1.5    1.5     1
0    0       0      0       0       0       0      0
0    0       0      0       0       0       0      0
0    0       0      0       0       0       0      0

Note that this table works fine assuming a cold/warm start & drive off under MOST conditions. If you get yourself stuck in traffic, or for the first several minutes following a heat soaked hot start where IATs typically soar, these numbers result in charge temps that are significantly lower than reality, at least until you start getting some airflow going to relieve the heat soak condition.

[hugemikeyd]

thank you so much for sharing, i was finding a similar shape and low numbers like you have, mine are just higher. I am going to try to lower mine and see what happens. I was finding that the engine would stall at idle when things got too hot.