tuning at altitude

GraemeD · Post by **GraemeD** » Sat Jan 05, 2013 7:55 am

quick question, I keep bouncing this back and forth. this is not a Baro correction question,

if you tuned an engine at sea level (100kpa) and set the WOT AFR to say 12.5. and had your 70kpa AFR at say 13.0.
Now take that same engine to altitude and at WOT you map is 70kpa, what AFR should you be running?

I guess the question is what AFR will give the best torque at 70kpa map? and would you want that same AFR at both conditions? (WOT at 70kpa baro and 70% TPS at 100kpa baro)

E4ODnut · Post by **E4ODnut** » Sat Jan 05, 2013 8:18 am

With no baro correction, assuming that best power at a given RPM for your engine WOT 100 KPA is at 12.5:1 AFR. The best power at the same RPM at altitude where WOT is 70 KPA should still be at 12.5:1 AFR, or very close to it. You will be making less power, and therefore less torque because with less manifold pressure because you cannot get the same amount of charge into the cylinders. Optimum AFR is not a function of manifold pressure.

subwoofer · Post by **subwoofer** » Sat Jan 05, 2013 10:04 am

GraemeD wrote: I guess the question is what AFR will give the best torque at 70kpa map? and would you want that same AFR at both conditions? (WOT at 70kpa baro and 70% TPS at 100kpa baro)

AFAIK the best torque AFR is independent of manifold pressure, so at WOT you want as much output as can be found. But for your example, I would argue that the 70kPa at 70% TPS should be run at max efficiency AFR while the same MAP at WOT means max available output.

As long as you can get more manifold pressure, there is no need to move away from max efficiency since you can get more power easily. All in all, I think this suggests that the AFR table should always be alpha-N, regardless of fueling mode.

wmax351 · Post by **wmax351** » Tue Jan 08, 2013 8:09 am

I would think %Baro would be a good option here. I am planning to switch my bike over to that. That is especially true with ITB's on the bike, where small throttle movements make a big change in MAP.

If using Incorporate target AFR, the VE table is just a VE table, so the % baro will be fine for the AFR tables

GraemeD · Post by **GraemeD** » Tue Jan 08, 2013 1:51 pm

I have my AFR table load set to %baro, but "Incorporate target AFR" is not turned on. so only when I have EGO correction on will it actualy do anything. I need to read some more on "Incorporate target AFR", my uneducated worry is the lag in the EGO sensor reading.

subwoofer · Post by **subwoofer** » Tue Jan 08, 2013 2:48 pm

Incorporate AFR and EGO are orthogonal, one does not influence the other.

TheSilverBuick · Post by **TheSilverBuick** » Fri Jan 11, 2013 4:31 pm

I live at 6500 ft elevation and regularly drive to sea level, and use the %Baro. Got my table squared away at home, then used the baro table to set the AFR's back to where I wanted them as I drove off the mountain, like every 5 or so kPa drop in baro. I've never used the AFR targets or EGO corrections.

sedd · Post by **sedd** » Sat Mar 01, 2014 12:55 pm

Could you respond back with the numbers you came up with for the baro correction? Are you using standard map and speed density to run your engine?

I think I read from other posts, that as you and others went from higher elevations to lower elevations you will have to lean out the pulse width? It is running rich so you need to apply negative values or values less than 100, depending on your tune tables?

The computer uses the same VE value from the tables based on map and rpm. Did you see the air fuel ratio richen as you went lower? If that is true then the engine must not be breathing as well (cylinder filling is not as good) and not getting as much oxygen in it for the same VE table settings?

Put another way, as you climb in elevation, the reduced barometric pressure improves the breathing of the engine and it will need more fuel and larger pulse width?

Attached is my current msq that I want to change now, just before retuning due to injector setting changes. Any suggestions on numbers to enter would be appreciated.

sedd · Post by **sedd** » Sat Mar 01, 2014 12:58 pm

this time with msq attached?

ol boy · Post by **ol boy** » Sat Mar 01, 2014 5:05 pm

Sedd.. I've seen quite the opposite.. reduce in elevation requires larger PW. I leave EGO on and check it while climbing long grades. I'm pulling 12% out at 6200 feet compared to 2000 feet.

To make baro correction work properly your injector dead time needs to be spot on and the VE table needs to provide a steady AFR throughout.

kaeman · Post by **kaeman** » Tue Mar 11, 2014 10:12 pm

hey SEDD, just like in the old days of carbs.... if you were a valley racer and went to play with the boys in the hills you always had to jet down or the engine ran so rich it made no power. Higher elevation lower pressure helping fill the cylinder, there fore less o2 and less required fuel..... lucky the you can use baro correction to help with changing elevation.

sedd · Post by **sedd** » Wed Mar 12, 2014 7:28 pm

I agree that the overall engine capability has to be lower at higher elevations. The map signal at wide open throttle is limited to the barometric map reading.

Possibly I have a misunderstanding of what is going on. If I understand how the VE table works, it looks at engine MAP and rpm and then sets the amount of fuel to spray based on that. Other corrections not being discussed. So for a given map and rpm, a person would expect the engine to pull in the same amount of air? The way I think, the barometric pressure has nothing to do with the map signal at an rpm, that map and rpm are not changing. As stated above, to me the barometric levels only restrict the upper levels of the wide open throttle potential. So why does a given MAP signal not have the same VE and a lower VE as you go up in altitude?

This is a very crude explanation. The SAE paper I noted is much better at the explanation. Apparently it has to do with the exhaust pressure being lower as you go up in altitude. That affects the cylinder pressure at the top of the exhaust stroke. There is an effect for both the exhaust valve and the intake valve. The exhaust valve is not as strong an effect as the intake, so the intake dominates. One cancels out the other but not in the same proportions, as I have noted. So the effect is to reduce the VE but not for the reasons most are thinking.

What is still open for debate is the effect of the air cooling as it passes the throttle plates. I have found data to confirm what others have posted about the temperatures getting very cold. But, it appears the air heats up again about as fast as it cools down. Another very crude explanation is that the temperature rise will match the velocity of the air. The velocity and temperature are very fast and cold at the throttle tips, and then as the air slows as it enters the intake, it will rise up again in temps.

There is a physics law that may apply, that says the enthalpy stays the same across a throttle device. If that is true the temperature change after it returns to the same enthalpy is less than a degree. I was able to find an enthalpy versus entropy diagram for dry air to check on this. The air first drops at constant entropy and then rises at a constant pressure line to the same enthalpy. An yes, this sounds way to much like a physics or thermo class.

Keep in mind that we are looking at the change in pressure drops and not final MAP pressure. The MAP pressure is the same. How would that change in pressure drop affect the temperatures, if any.

What I am not certain about is the effect of humidity on this process. If the water vapor in the air turns into liquid fog or ice crystals, that may delay the rise of temp. I was surprised that many car mfg add heaters to the throttle bodies to combat ice formation. Heat is removed from the air and is tied up in the water, it might take some time to convert that energy back into the air. Could the air travel fast enough with the water as a fog and thus be more dense as it enters the cylinder? The same type of diagram for the moisture does not show the low pressure areas well and I don't know for certain what temperatures the water vapor has after the constant enthalpy process. I did find a set of psychometric charts for various elevations and that shows at constant enthalpy there is almost no temperature variation. Eventually I think the air and water vapor will rise in temps and the water will return to vapor. So what else? maybe that as you rise in elevation the amount of humidity tends to be less. People who live at elevation can correct me on this. Temperatures are also lower but we have assumed that this is taken care of by the MAT/IAT sensor.

TheSilverBuick · Post by **TheSilverBuick** » Tue Mar 25, 2014 9:13 pm

I run my fuel on Baro% and my spark on standard Speed Density. The high'ish compression of my Buick's engine likes the Speed Density on spark since it will limit advance in denser air and keep it from pinging (I use the MAT timing retard as well, Las Vegas heat in summer can be brutal).

It definitely LEANS out going down the mountain without any corrections, which conversely richens up going up hill without corrections. Even my carbureted cars that were tuned essentially at sea level run rich at high elevation. I tune my correction table using actual conditions. Driving at 70mph, at a fixed Load (%Baro) and RPM, my AFR should be 16:1. As I drive off the mountain I usually wait until I get a bit of lean surge or mis-fire, then pop the laptop open move the dot on the Baro Correction table upwards until the AFR is back to 16:1. The safest way to do this is to actually stop and bench mark and re-calibrate it at idle.

You are really over and under analyzing this. Unless you are an air flow physicist the calculations are complex. Talking temperature delta's across the throttle blades without taking into account the air density drop (vacuum), which if you drop the density of a gas it acts the same as expanding a gas which is a cooling effect. VE's change primarily from the exhaust side because the intake is already under a vacuum in all part throttle driving and less impacted. Then as you noted there are a dozen other variables that impact the VE. Many people have a hard time getting that air pressure/density and O2 concentration are separate things.

kaeman · Post by **kaeman** » Sat Mar 29, 2014 7:55 pm

just remember that the change in altitude gives that same kind of effect as the change in weather, on a overcast day the baro pressure is lower, and a hot clear day has higher baro... but the engine still sees 70 kpa as 70 kpa, doesn't matter if the change in altitude or the weather is making the difference, unless you are using baro corrections the engine will still try to hit the ve target for whatever kpa and rpm bins it is using for running.

if you are using the constant baro correction ie 2nd map sensor then the altitude and weather will be taken into effect always, otherwise it will be determined at each startup and not taken into account until restarted.

sedd · Post by **sedd** » Sat Mar 29, 2014 8:51 pm

I am not a physicist, I have an engineering degree. I did compare the air at both atmospheric and vacuum conditions using a chart I found. There is a common thought that the air cools and this affect the air entering the cylinder, but I didn't find any data or theory that proves this. There is only a cooling effect from the throttle blades, but I feel this occurs only very close to the blades. I don't see cooling of the air charge as it enters the cylinders taking place. Based on the comparison and research I did the air will have heated back up as soon as it slows down in velocity.

I feel I have eliminated that as a factor in why density changes affect the AFR. I don't think the density of the air outside the vehicle or the humidity of the air will make any significant change in AFR.

I still would like to see some data from folks who have good correction curves, there is not a good way to prove the theory for the spreadsheet I developed without that data.

pit_celica · Post by **pit_celica** » Sun Mar 30, 2014 7:05 pm

I understand your confusion about the point : "The same MAP/RPM combination should yield the same ingested air (hence the same fueling requirement to keep the same AFR), no matter the barometric pressure".

But, you need to understand that the problem do not comes from the MAP value, but from the VE value. In a speed density system, the real meaning of VE is "how much air mass is getting inside the cylinder vs the maximal air mass that can be stored inside the cylinder at ambient conditions". Mathematically, the VE can be calculated this way : VE = actual_mass_of_air_inside_cylinder / (ambient_air_density * cylinder_volume).

The first term (actual_mass_of_air_inside_cylinder) is calculated with the MAP, MAT and RPM values. The cylinder volume is fixed for a given engine. The only remaining variable is the ambient_air_density. If we say that the ambient temp is fixed, then the only other variable affecting ambient_air_density is the barometric pressure.

The problem is that when you tune a VE table using RPM as the X axis and MAP as the Y axis, there is no provision to compensate for the barometric pressure. This is why there is a GBaro term in the fueling equation.

This means that an engine at sea level have a theorical VE value higher than the same engine running at the same MAP/RPM/MAT conditions at higher altitude. Because your Megasquirt used a MAP/RPM VE table, you need to compensate for the change in barometric pressure (change in ambient air density) to the VE value. As you may already know, a lower VE value require less fuel to keep the same AFR. This is why you need to remove fuel when going in higher altitude.

Sam

sedd · Post by **sedd** » Wed Apr 02, 2014 7:51 pm

I would have thought the VE would have been based on a "standard" atmospheric condition and density only, and would not be based on "ambient" air density.

I thought the VE tables were based at standard conditions are adjusted to correct for non standard conditions, ie temperature. pit_celica, are you certain about your statement as it applies to MS3 program?

Can we agree to this? Let's assume a temperature and back pressure on the exhaust don't change. If the pressure underneath the throttle plate, and in the intake manifold is the same, the engine will get the same amount of mass of air into the cylinders and not depend on the barometric pressure.

pit_celica · Post by **pit_celica** » Thu Apr 03, 2014 9:32 am

sedd wrote: Can we agree to this? Let's assume a temperature and back pressure on the exhaust don't change. If the pressure underneath the throttle plate, and in the intake manifold is the same, the engine will get the same amount of mass of air into the cylinders and not depend on the barometric pressure.

I do not agree. Remember that the throttle plate act as a a pressure drop. PressureInlet - PressureDrop = PressureManifold. If the PressureManifold is the same, but with a lower PressureInlet (lower Barometric pressure), then the PressureDrop need to be lower to satisfy the equation (the throttle need to be more open, less restrictive). Also, because the pressure drop is smaller, this means that the flow will also be smaller (a fluid moves because there is a delta_pressure applied on it).

All this to say that at high altitude, you will need to ingest more air (more volume, throttle more opened) to have the same load on the engine (same MAP at sea level).

Does it all make sense? In my head, it's logic, but I'm open to discussion.

Sam

sedd · Post by **sedd** » Thu Apr 03, 2014 5:45 pm

I appreciate the dialog. I will not likely drive much at elevation, so my interest is purely for the fun of the theory.

I agree to a point. The throttle plate will have to open a bit more to achieve the SAME pressure of MAP.

Since the pressure below the throttle plate is the same, the density of the air, and mass of O2 in it should be the same. Pressure drop is what creates the flow across both the throttle plate and across the valves. The flow across the intake valves should be relatively the same because the pressure = MAP in the intake manifold is the same.

The confusion is about why for the same MAP and RPM on the VE table, the pulse width of fuel does not give the correct air fuel ratio. The baro change affects the air flow in some way, and so far nobody on this site appears to have found an equation to describe what is going on.

My research found an article that described how the exhaust pressure change due to baro affected both the intake flow and the exhaust flow. So that article notes there is an effect due to pressure change, but not from the pressure on the intake manifold side.

I have not got any response from those who actually drive at various elevations. I gathered a small amount of info from various posts and places them on a chart. So far the equation from my research appears to be a bit too linear to match up with real life, so that research is close but might not be the final answer.

TheSilverBuick · Post by **TheSilverBuick** » Sat Apr 05, 2014 9:05 am

For a little follow up. My trip off the mountain, the barometric reading is right above the correction chart. I checked two AFR/load points to make sure the correction was where I wanted it. Cruise was at 19:1, which to me was not as repeatable as I'd like, then 13:1 with just a little bit of acceleration, basically 60% load at 3,000rpm on the %Baro. This is my Firebird that pulls 16inHg vacuum at 650rpm idle (at sea level) and seems to be A LOT more tolerant of lean conditions than my Buick that pulls about 10inHg vacuum at 900rpm idle (at sea level). My Buick I run the highway at 16:1 and it starts mis-firing going leaner so bench marks much cleaner, but this is a test/calibration of what I was running that day. The EGO corrections are turned off (hence the 100%). Also, this is the release version of MS3 v1.3.0.

Baseline reading. 78-82 kPa is the normal baro range here.

Then about here I started cursing myself as apparently the baro correction is capped at 120%.... Pretty much at this point, I started raising the Required Fuel (for right or wrong..) on the fly to keep from going too lean.

Then from essentially sea level for my trip home, I raised the Required fuel number more then globally lowered each point of the correction curve to get me back to the proper AFR and have some adjustment room at the top. This is the table I ended with and is what works on my car at least down to 80kPa barometric pressure. The "curve" is waaaay linear than my previous curves, and I believe it's because my Total Vacuum % and Rate% are both zero's now and before were the default 147/-47, which were numbers I didn't understand and was likely making corrections to the corrections that were giving my old correction table a cubic curve from my standard zero correction point. Or so is my running theory. I've repeated the above test/calibration multiple times since 2009, and this is the first time I've gotten a linear curve and my Skylark, when the 147/-47 were set to zero had all sorts of tuning weirdness happen when I was driving up to 65kPa (11,000ft) last September that it didn't have before, aka I found myself adjusting a baro correction curve that had been good for the previous two years.

The test subject.

Typical view for nearly 250 miles of the 500 miles I drove each way. Hence, as long as I can keep it between the rumble strips, I tune on the fly. Ain't much to hit.

Averaged 20mpg with 4.56 gears, 2,800-3,000rpm cruise speed at 70-75mph with LOTS of mountains and 6% grades running several miles at a time.

Megasquirt Support Forum (MSEXTRA)