To disable cooling flaps or not, The endless debate Options

[Superhawk996]

Arduinos are cheap and the selection of sensors are pretty good. No data storage issues either.

Good point but I’m a lazy bastard and don’t want to have to build my own Data Acquisition out of amplifier modules (Arduino can’t read a thermocouple directly), SD card shield, etc., and have to do the programming too.

But it’s honestly not a bad idea vs used DAQ for thousands of dollars.

[JamesM]

wondering if there are things I can implement to increase temps under cold decel situations as it looks like even with the flaps the cooling is pretty pronounced.

I would think you could get something by running a lean burn and advance timing short of decel fuel shutoff that might help but may be difficult to tune without knock sensors.

That data that I posted is ancient - that book was 1st published in 60’s so I would almost guarantee that was a carbureted engine.

Lean burn with advanced timing is how my heads are running 75 deg cooler than stock at cruise as it is (IMG:style_emoticons/default/biggrin.gif)

Head temps drop off quicker on the lean side of stoich than they do on the rich side. Added bonus I get near 40mpg and my oil stays way cleaner too, which itself should be better for engine wear.

Around stoich is the hottest but limiting that to only during conditions where you are experiencing extreme cooling would be the challenge.

Driving down a mountain in freezing weather is a bit of an edge case for me given they salt the roads here in winter so for now ill take the trade off, but its an interesting problem to think about.

[Superhawk996]

its an interesting problem to think about.

Indeed. I would have guessed you would be running over on the rich side for power.

In a lot of ways air cooling is simply archaic and has some pretty extreme limitations. The downhill grade in cold weather is going to be one of them.

But honestly that’s part of the “charm” that comes along with the simplicity of air cooling.

And since all real Porsche’s are air cooled, I’ll live with them. (IMG:style_emoticons/default/lol-2.gif) (IMG:style_emoticons/default/aktion035.gif)

[Nogoodwithusernames]

What does your tune look like? I followed John Connolly's advice/thread over on the Samba for trying to tune my Megasquirted 1.7 T4 (In my Squarback) for lean cruise and could never get head temps down under 350F, running 12.x at cruise now and heads are 320-340F at 60-70MPH. Varies a bit with ambient temp. I only get 19-21mpg with mixed driving though.

I do run a stock working Tstat in both the 914 and the Squareback, with all tins (including custom sled tins to keep it out of ambient airstream on the VW), but I've never checked position at various loads or temperatures on either so I don't have much to add in that regard. I will say I don't drive the 914 when it's 90+ outside because that's miserable without AC, much preferable when it's 60-70F outside.

wondering if there are things I can implement to increase temps under cold decel situations as it looks like even with the flaps the cooling is pretty pronounced.

I would think you could get something by running a lean burn and advance timing short of decel fuel shutoff that might help but may be difficult to tune without knock sensors.

That data that I posted is ancient - that book was 1st published in 60’s so I would almost guarantee that was a carbureted engine.

Lean burn with advanced timing is how my heads are running 75 deg cooler than stock at cruise as it is (IMG:style_emoticons/default/biggrin.gif)

Head temps drop off quicker on the lean side of stoich than they do on the rich side. Added bonus I get near 40mpg and my oil stays way cleaner too, which itself should be better for engine wear.

Around stoich is the hottest but limiting that to only during conditions where you are experiencing extreme cooling would be the challenge.

Driving down a mountain in freezing weather is a bit of an edge case for me given they salt the roads here in winter so for now ill take the trade off, but its an interesting problem to think about.

[930cabman]

Wondering if the boys (girls) in the fatherland went after this?

(IMG:style_emoticons/default/screwy.gif)

Lots of numbers for the data guys

[Root_Werks]

I'm one of those weird people that installed a functioning T-stat system on my Bug (didn't have one) and installed a T-Stat on my wife's 72' Bronco which didn't have one. Waiting an eternity for an engine to warm up on cool/cold PNW days straight up sucks. Regulated cooling systems rock.

(IMG:style_emoticons/default/aktion035.gif)

[930cabman]

I'm one of those weird people that installed a functioning T-stat system on my Bug (didn't have one) and installed a T-Stat on my wife's 72' Bronco which didn't have one. Waiting an eternity for an engine to warm up on cool/cold PNW days straight up sucks. Regulated cooling systems rock.

(IMG:style_emoticons/default/aktion035.gif)

Don't feel alone, all (or at least most) of us here are weird (IMG:style_emoticons/default/hissyfit.gif)

[JamesM]

its an interesting problem to think about.

Indeed. I would have guessed you would be running over on the rich side for power.

Well... I am. At least in the areas where I need to be. A beautiful feature of programable EFI is I can map things in a non linear fashion.

75% load and above i am running ~12.8:1 below that I am running ~15.5:1


The ideal AFR for maximum power is dependent on the ignition advance, and vice versa.

Flame front speed and thus the timing of max cylinder pressure changes with AFR, so as long as you advance your ignition timing to account for the slower flame front you don't wind up loosing anything (under part throttle conditions) other than unnecessary fuel waste and heat. I actually find part throttle to be more responsive/powerful with this tuning vs how it was with d-jet and find less throttle angle is required to maintain the same speed.


More power, cooler heads, cleaner oil, and significantly better efficiency than stock (IMG:style_emoticons/default/biggrin.gif)

My apologies for veering off topic a bit here.

[dr914@autoatlanta.com]

never

[Jack Standz]

As has been said before, this is awesome! "How works a vw thermostat (long read, data heavy)":

https://www.thesamba.com/vw/forum/viewtopic.php?t=705328

From the guy who makes replacement t-stats for 914s/late 912s as well as other aircooled motors (VWs).

[Superhawk996]

its an interesting problem to think about.

Indeed. I would have guessed you would be running over on the rich side for power.

Well... I am. At least in the areas where I need to be. A beautiful feature of programable EFI is I can map things in a non linear fashion.

75% load and above i am running ~12.8:1 below that I am running ~15.5:1


The ideal AFR for maximum power is dependent on the ignition advance, and vice versa.

Flame front speed and thus the timing of max cylinder pressure changes with AFR, so as long as you advance your ignition timing to account for the slower flame front you don't wind up loosing anything (under part throttle conditions) other than unnecessary fuel waste and heat. I actually find part throttle to be more responsive/powerful with this tuning vs how it was with d-jet and find less throttle angle is required to maintain the same speed.


More power, cooler heads, cleaner oil, and significantly better efficiency than stock (IMG:style_emoticons/default/biggrin.gif)

My apologies for veering off topic a bit here.

No I think this is an awesome discussion about how to manage temperatures and the thermostat and flaps are only part of it.

The Mackerle book was representative of the state of air cooling at the time the 914 was developed. Similarly the factory D jet tuning that was available at the time has some huge limitations that EFI can easily overcome. Your edge case decent and learn burn tuning at light throttle cruise being prime examples.

Right on target of why I brought this discussion out of Ron’s thread.

What could be super cool would be to use some stepper motor control or PWM solenoid control to better optimize the flaps position along with EFI but would need a good failsafe strategy that fails flaps open like the OEM thermostat.

[emerygt350]

As has been said before, this is awesome! "How works a vw thermostat (long read, data heavy)":

https://www.thesamba.com/vw/forum/viewtopic.php?t=705328

From the guy who makes replacement t-stats for 914s/late 912s as well as other aircooled motors (VWs).

That is great stuff.

[JamesM]

What does your tune look like? I followed John Connolly's advice/thread over on the Samba for trying to tune my Megasquirted 1.7 T4 (In my Squarback) for lean cruise and could never get head temps down under 350F, running 12.x at cruise now and heads are 320-340F at 60-70MPH. Varies a bit with ambient temp. I only get 19-21mpg with mixed driving though.

Few questions

What spark plugs (part number) are you running?
What cam?
What AFR were you running on lean cruise?
What ignition advance on cruise?


the squareback is a little more of a brick so I would expect head temps to possibly be a little higher than a 914, but my 2.0 test car (MS converted bone stock 2.0) never saw over 350 when I took it on its maiden voyage to California and back. 37 MPG on the stretch back between Barstow CA to Mesquite NA, was driving conservatively to see what I could get, so 65-70 MPH. Had a couple outer wheel bearings failing at the time as well along with a corner balance issue, not to mention most of the route was going up in elevation so I suspect I may be able to do even better.

I got 23mpg on my last camping trip in my 1.9 wbx 85 vanagon running basically the same system with a very similar tune though, so I would think for sure you should be getting better than 20 in the square as nothing is more of a brick than a vanagon loaded with camping gear.

Not sure what other details I can give about my tune, 12.8 @ 28 deg at max load, leaning slightly to low 13s around 75%, then a very abrupt transition with everything below 75% targeting 15.5 @40 degrees advance. You want to make sure NOTHING is producing AFRs between roughly 13.8 and maybe 15 as that is where you are going to be hotest. the further you can stay either side of that range the better from a heat standpoint. Idle is around 13.5-13.8 with idle speed being stabilized by timing control (NO IAC) TB opened up enough to require the timing to be in the 0 to 5 retarded range. Some custom shaping in the higher vaccum areas around 2000 rpm to ensure everything drops back to idle ok. This is on a 2.0, on a 1.7 that might need some minor tweaks given the slightly higher compression and smaller bore.

when I initially put the plug and play MS system together my initial thought was to have a d-jet mimicking tune and then my own "Special" tune for people to pick from. I roughed both of them and and drove around on each, the "special" tune was so much better in every way I never went back and fully refined the d-jet mimicking tune. I probably will eventually just to have it, but I see little point to it now.

[brant]

Full load redline is the most important for mixture

Something around 12.5 or close to that is rich enough to increase cooling on the heads and ideal for max power

Leaner usually raises head temps.


Also idle doesn’t matter
Cruise barely
AFR becomes important at high rpm
Race cars are jetted for red line

[JamesM]

Ok, you said I was still on topic so... here goes, you asked for it (IMG:style_emoticons/default/biggrin.gif)

Leaner usually raises head temps.

This is what you generally read on the internet but the reality is a little less one dimensional and to be a little more scientific you need to give a frame of reference.

Specifically leaner than what?

Head temperature response to AFR is not a straight line, it is a bell curve. Peak head temperatures typically occur near stoichiometric mixtures, with temperatures falling as the mixture becomes either richer or leaner. So say you are running at 12.5:1 at a given load, it is true that running leaner say 13.5 or 14.5 will produce hotter temps BUT that is only true to a point. If your starting point of reference is ~14.7:1 and you run leaner, things get cooler.

The primary source of heat in a combustion engine is the chemical energy released during combustion. Maximum combustion temperatures generally occur near stoichiometric mixtures (around 14.7 AFR). To keep heads cooler you either need to increase cooling OR put less heat into them to start with. Running richer than 14.7 you are still roughly burning the same amount of fuel, now just with a little left over (fuel rich) Because nearly the same amount of fuel is still being burned, the total chemical energy released changes very little. The excess fuel acts as a heat sink through vaporization and by absorbing energy that would otherwise increase combustion and component temperatures.

The other way to get less heat in the system, is to generate less heat to begin with. 14.7(give or take) is stoichiometric, the maximum amount of fuel that can be completely combusted. Burning a leaner mixture than that generally results in less total heat being released into the engine. From a cooling standpoint its more efficient to not generate the heat in the first place than it is to put the heat in and then cool it off.

Now I know at this point I have raised a LOT of questions, but hang on because there is a lot more to explain here that may answer them...

Most car guys on the internet only care about power and when they are talking about rich vs lean its always in the unmentioned context of full load/wide open throttle. In this context, they are correct, leaning out your mixture would most likely be very undesirable for multiple reasons. Heat, potential detonation, but possibly the most important in the context of full throttle, POWER.

Going richer than stoichiometric you will notice power output vs AFR does not change much largely in part due to the fact that you are not extracting any more energy from the gasoline the more you add. On the lean side of 14.7 however power drops off FAST, so if you lean burn under high load, assuming your engine doesn't detonate, you are significantly down on power under the operating conditions (full throttle) where you are looking for maximum power.

But under conditions where you do not need the power (Cruise, decel, light acceleration you can get away with not burning the extra fuel and it is in fact preferable because you are 1. saving fuel and 2. generating less heat.

BUT in order to do it there are other things that have to be considered, specifically flame speed. Say for example your engine is tuned for max power and at a specific RPM, say 3500 that happens to be from 12.8:1 AFR @27 degrees advance. This happens to be the AFR where you are typically going to have the fastest flame front. Assume a centrifugal advance distributor so your timing at 3500 doesn't change (27 degrees) you will lose power going leaner OR richer from here but initially its not due to combusting less gasoline because anything richer than 14.7 you are already utilizing essentially all of the available oxygen for combustion. Instead the power output changes because the change in flame speed means the peak cylinder pressure is no longer occurring at the optimal time. Then once you go leaner than stoichiometric factors contributing to power loss are twofold 1. you are now starting to burn less fuel than the available oxygen allows for, so less energy into the system and 2. The flame front continues to get progressively slower meaning the fuel you are burning is ALSO creating sub optimal cylinder pressure. You will generally find that with a stock distributor/timing curve you can be pretty much anywhere between 14-11:1 afr without noticing much, but things become undriveable pretty quickly as you go leaner than 14.7. You will most likely be bucking and on the edge of even having enough power to go down the road, popping out the exhaust as the flame front is slow enough that fuel is still in the process of combusting by the time the exhaust valve opens.

So the actual drivability problem of trying to run leaner than 14.7 under light load conditions (which in a street driven 914 is like 90%+ of your driving) isn't that there is not enough energy in the amount of fuel, its that the energy you are producing is all flying out your tailpipe because you are unable to optimize when peak cylinder pressure is occurring. This can be solved however by igniting the fuel sooner and is a fairly easy problem to solve with a programable three dimensional timing map.

The short take away here, to generate the necessary power needed at light loads you can either burn rich, throwing more fuel than then you have available air to burn and removing some excess heat with the extra fuel OR add advance to account for a slower flame front and burn less fuel than the available air


Congratulations if you made it this far!

Its a multidimensional problem and takes a little while to wrap your head around.

Ive been playing with EFI on 914s for 20+ years now and its only been in the last couple where I have really started to nail down some of this stuff. You will find older posts where I say you cant run lean, it was because I didn't know better.

[emerygt350]

Great, now I need a new efi....

[930cabman]

Go James Go

I'm sure it's all good stuff, but for me I will check a spark plug if things don't seem right.

[Superhawk996]

As has been said before, this is awesome! "How works a vw thermostat (long read, data heavy)":

https://www.thesamba.com/vw/forum/viewtopic.php?t=705328

From the guy who makes replacement t-stats for 914s/late 912s as well as other aircooled motors (VWs).

That is great stuff.

Hadn’t seen that before. Good stuff.

Hilarious (but not unexpected) that people over at Samba going
Through the same debate. Hats off to the Awesome Powdercoat guy for leading the charge.

The more I think about it, the more I’d really like to do a full blown data acquisition on a T4 someday. When I first started my career strip chart recorders were still a thing which really limited the scope of data acquisition. That was in the early 90’s. By the mid 90’s with the take over of PC and hard drive based DAQ systems, sensors becoming smaller and smaller, dynos and development trips were soon running up to 300 channels. I can hardly imagine how limited data acquisition would have been in the 60s. The Mackere book sort of reflects this. Although it’s an engineering textbook, with select data related to the topic being discussed, it’s clear they had to work a whole lot harder to get the data I would now take for granted.

[rudedude]

@jamesm
Would you share your microsquirt tables?. I’d like to try your lean burn on my 2056. I’m curious what they look like.

[73-914]

I'm one of those weird people that installed a functioning T-stat system on my Bug (didn't have one) and installed a T-Stat on my wife's 72' Bronco which didn't have one. Waiting an eternity for an engine to warm up on cool/cold PNW days straight up sucks. Regulated cooling systems rock.

(IMG:style_emoticons/default/aktion035.gif)

The Bronco didn't have any thermostat?
Good way for it to overheat