LSPI A KNOCK ON THE NEWEST ENGINES 2/4/16

[Kjewer1]

The main thing I'd be worried about with the diesel oil is that the problem additives are allegedly detergents, and diesel oils amy tend to be high detergent to deal with the soot problems. Just a thought, I'm no expert on motor erl.

Sponsored

 

[JWC]

Are there any fuel brands that are preferred for maximum protection? Is there really that much of a difference between brands?

 

[arghx7]

Watch your boost at highway speeds when you climb hills. You can get it to 10-15 psi at very low throttle angle. Flooring it, punching it, etc has nothing to do with LSPI. Doing so causes a downshift (on autos) and that's good, since it raises RPM out of the danger zone. On a manual that's up to the driver I guess.

15 psi boost is likely to be around 276 ft lbs using 4 ftlbs per psia per liter as a conservative estimate, and that puts BMEP around 20 bar. ;) That's a ton of BMEP when you consider the low RPM where these engines can generate those pressures, and well into the accepted LSPI range. This can happen during perfectly normal easy driving with these cars.

I think we can agree that more load at low speed = higher risk for LSPI. I don't agree that boost in part throttle (defined here as not having the pedal all the way down) hill climbing is such a great metric to understand that though. The reason for this is that the VCT moves the cams in such a way that it requires a lot of boost to make torque, but that's done intentionally for the purpose of fuel economy. There's a much more detailed way to explain it but I'll skip it.

Requested engine torque will tell you more (need an AP to get that).

 

[Kjewer1]

So you're saying that the manifold pressure can't be used to calculate BMEP because of cam timing?

 

[Kjewer1]

I just checked requested torque at ~1750 rpm, 10.75 psi, part throttle cruising on the highway in 6th gear. The requested torque on the AP was much higher than my estimates in this thread, but about 100 ft lbs. 329, vs about 230. That would be a BMEP of ~23 bar, which I find hard to believe.

Calculating torque by airflow and rpm it comes out much lower, around 100 ft lbs, so you may still be on to something. I can collect some different data and calculate VE. It should be pretty bad if the cam timing is expected to limit BMEP.

 

[Kjewer1]

Got a quick log on the way home. 2166 rpm, ~9 psi boost, 28% throttle. The math says that VE comes in at 97%. Not that bad. Calculating torque by airflow it's 270 ft lbs. Using the standard 5 ftlbs per psia per liter it's 276 ft lbs. That's a BMEP of ~19 bar. Requested torque this time was 250, much closer. That's a BEMP of ~18. Not much difference there, and that's still awfully high for only 9 psi and 28% throttle.

I still think it's safe to say there's no need to floor it to be at risk of LSPI. The only real unknown at this point is what BMEP really starts the threshold for LSPI. According to most documentation it seems to be around 14 bar, but that's not specific to the ford engine.

 

[arghx7]

So you're saying that the manifold pressure can't be used to calculate BMEP because of cam timing?

You're basically asking how to calculate torque. The answer is that you can't do it very accurately. The stock ECU can somewhat do it (within 10% ish usually), on a stock car, but it's very difficult to be accurate with some simple formula in Excel or on paper. I don't know if the AP shows the modeled torque or not, I'd have to dig through the parameter list.

BMEP is the same as brake torque (flywheel torque), just with applying a formula to convert it to pressure. To get BMEP you need the indicated torque, the pumping losses, and the frictional losses. There are characteristic curves in the stock ECU for each, as mentioned in the diagram above which is basically for Net Mean Effective pressure (indicated - pumping).

Got a quick log on the way home. 2166 rpm, ~9 psi boost, 28% throttle. The math says that VE comes in at 97%. Not that bad. Calculating torque by airflow it's 270 ft lbs. Using the standard 5 ftlbs per psia per liter it's 276 ft lbs. That's a BMEP of ~19 bar. Requested torque this time was 250, much closer. That's a BEMP of ~18. Not much difference there, and that's still awfully high for only 9 psi and 28% throttle.

I don't know how you're calculating torque by airflow but you have to account for residual gases (VCT exhaust cam timing), manifold filling (VCT intake cam timing), spark efficiency, lambda efficiency. Pressure drop across the throttle valve is only one component.

Rules of thumb from muscle car magazines and sticky threads for 90s imports are just a waste of time for calculating brake torque. You have to account for friction as well - a low friction engine doesn't work as hard (lower IMEP) to make torque as a high friction engine.

 

[Kjewer1]

Torque by airflow is just the assumption that airflow is horsepower, and with rpm, you can get torque. Airflow is easy enough to get. The ECU provides it's calculation, which can be corrected against the WBO2, or you can calculate airflow yourself from the WBO2 and fuel flow. Airflow at 100% VE is certainly easy to calculate. Airflow equals hp is a big assumption, but anyone with a dyno can get the actual number for both.

I'm really not that interested in calculating torque (the diagram you provided certainly makes sense), but for some quick and easy calculations to get into the ballpark, it'll have to do.

In the end, I think I'm just trying to mathematically prove what we've already seen. Engines at low rpm and low TPS grenading. The proof is out there. Dealers have pulled the data and this is what they have reported. The fact that the basic calculations suggest that BMEP is in the accepted LSPI range shouldn't be unexpected IMO.

Here's another thing to consider. What happens when you reach full turbo spool at ~50% throttle/low RPM, and then go WOT? Essentially nothing. So how do the two differ if RPM and load remain the same? I could try it with this car and see how the ECU changes ignition and cam timing, which I believe is the distinction you're trying to make here. In think in either case we'd find LSPI worthy MBEP levels.

Your point is certainly valid, and I appreciate it. Above all I appreciate the technical discussion.

Sponsored