It has been burned into my brain when I read it the first time that when going up a hill drop a gear (5th to 4th) to increase cooling thereby lowering the temps of the engine. Raby, Mark Henry, I'm sure others have stated the same thing.

I must admit my sample size is small (3 cars) AND 2 of which already had pre-existing issues with the engine health, but in all 3 scenarios when cruising on the highway going 75/80 dropping a gear down only momentarily lowers the temps only to shoot them back up possibly even higher as I am trying to maintain speed at a lower gear which requires more pedal. In each of the 3 cases I have been using a CHT over cyl 3. Two of which were a Dakota Digital and one was an aircraft spruce analog gauge. The one engine that was not plagued with less ponies was a rebuilt 2056 (by me). Maybe I built the engine wrong, I don't know. I guess I would like to hear what others have experienced. I am also not doubting the physics. More RPM = More fan turning = More cooling. Lowering gearing means the engine does not have to work as hard to pull up the hill. That last statement should have a caveat though. There must be a certain point of force vs gear ratio vs drag vs speed that would determine that staying in 5th would ultimately be more beneficial than dropping to 4th as far as work the engine is doing (probably a question for Phil - @superhawk996 )

So am I wrong in my limited testing and it does in fact shed heat driving on the highway 75-80 mph with ambient 85 - 90 or was it just a myth started because the physics seem to point that direction.

I noticed this too when I was fighting a poorly timed 2056. I am interested to hear others. I have found that sustained effort up steep slopes results in a 'peak' temp and it tends to stay at that temp (in lower gears). I bet 5th below 3k is just going to get hotter and hotter.

I’m going to state right upfront I don’t have an all the data in the engine operating and cooling characteristics to back my opinion.

Having driven with and observed all kinds of drivers ranging from friends and family to so called “professional” journalists, I’ll say without hesitation America is the land of Luggers. Most Americans simply won’t downshift a manual trans when they should. Must be something in the water.

What we do know as fact from physics:

Dropping a gear (or two) is the right thing to do - if for no other reason than to increase mechanical advantage that lower gear ratios provide to ease the load on the engine. Increase of fan speed might be a little bonus.

As far as the physics go, it’s going to take a certain amount of work (power) to climb a given hill. Regardless of what gear it’s done in, the work expended is the same.

For the aerodynamics, it will take a given amount of power to maintain a given speed. Power needed is proportional to the cube of speed. So let’s say you wanted to double speed from 30 mph to 60 mph, it will take 8 times as much power to maintain 60 mph as it did 30 mph.

So what is unknown is where the engine is operating at peak efficiency. If we have a dyno curve for a specific engine we would know. But what makes it difficult to known exactly where the sweet spot is in the real world.

To do the work of the climb, we want peak torque. But to maintain high speeds (highway speed) we want max horsepower. Not sure exactly where its sweet spot is for any random hill at any particular speed with a given engine.

Then with respect to engine cooling, so much of that is dependent on the operational tune of the engine - as stated timing advance is going to have an effect as will AFR. Running lots of advance, with lean mix, and high load is a death sentence for an engine.

Cooling is harder to pin down - air cooled engine cooling is governed primarily by the temperature difference between ambient air and the cylinder head temp. I recently posted a chart of the cooling coefficient of air vs air velocity. The bottom line is that it isn’t linear. Doubling air velocity from the fan doesn’t double the cooling effectiveness. So at some point, the increased fan speed gains less and less cooling increase. (Chart in link below)

http://www.914world.com/bbs2/index.php?sho...amp;pid=3097424

Here’s a part that is counter intuitive. The hotter the heads get, the more effectively they cool. But of course we can’t have heads at 600F or we run into other issues like dropping valve seats or stretching valves.

And ambient air temp - we can’t do much with that other than to recognize that if the cooling is dominated by the difference between heads and ambient air, the ambient temp isn’t the biggest contributor. Example, original baseline heads 350F. Ambient air 90F. Difference 260F.

Now assume cooler ambient air = 50F. Temp Difference 300F. So that change in cooler ambient air only increases cooling effectiveness by 15%. Not as big as you might have thought? We almost halved the ambient air temp but only got a 15% cooling increase.

Now let’s do the same thing with increasing head temps. Head 400F. Ambient air 90F. Difference 310F. That results in a 19% cooling increase vs baseline with 350F heads. But wait, we only increased the head temperature 14% (350 to 400F) but we got a 19% increase in cooling.

See why air cooling is a bit weird? Actually cools better when head temps have increased.


So back to the question - when is it better to down shift to find the sweet spot for optimum cooling? Who knows without some very complicated modeling or fully instrumented road testing

So to offer some sort of more definite conclusion, let me share this from years of instrumented testing - trailer towing at max GVW up grades like Davis Dam, Eisenhower, Grapevine, etc.

Since the power needed to overcome highway speed aerodynamic drag is large (remember - varies as cube of speed) - slowing down will free up power to make the climb.

There are situations in modern cars where the best that can be done is have the trans kick down gears and let the vehicle speed drop. I’ve been in many test vehicles and the fully loaded climb up a grade is made at 45 mph in the right lane - flashers on, engine running around 4000-4500 rpm, timing being retarded by FI, and coolant temp is hovering at 250F, oil and trans temps at 250F or more.

Simply can’t maintain highway speed overcoming aero drag and still have enough power to do the work required to climb the grade with a load.

914 is no different in that regard. If you can’t make the climb and maintain speed, just downshift (for mechanical advantage of the gearing) and let the speed drop a little. That is going to be the most beneficial path to avoiding excessive head temps.

Living in the mountains near Salt Lake I've had a lot of seat time in front of CHT and oil temp gauges. Downshifting definitely has positive effects on cooling. Lugging increases temps quickly. That said, there's definitely a sweet spot between gearing, RPMs and speed. For me, @superhawk996 s observations bear out...dropping down a gear and going a bit slower seems to have the most positive effect. And of course that relationship changes depending on grade, humidity, and a bunch of other variables. I will say this...having a CHT gauge has taught me a lot about how to drive an air cooled car. I also sometimes want to turn it off because ignorance is bliss. Maybe that's why they didn't come from the factory with them

Very interesting stuff SuperHawk. Thank you for chiming in on this.

I guess the definition of lugging is important. With the gearing the way that it is, I don't think running 3300-3500 up an incline is lugging. Probably wrong.

I do know that my latest test with an Ljet engine, the timing was spot on. I think the key to what was said was that it dropping a gear should come with dropping speed.

On a side note somewhat related. I did touch 425 on cyl 3. Valves were fine after. Not saying it is a good thing, just saying my valves did not stretch or change after that Hi-heat beating. More so an observation.

Also, ignorance is bliss. Not having the cht, I don't know if killing cyl 3 or not.

Just as a point of reference, in aircraft engines (air cooled Al/iron sleeved cylinders and Al heads) Lycoming specifies a max never exceed head temp of 500 F and a max continuos of 435F

Most prudent operators prefer 400F max continuos.

I went digging the other day for a 'real' value on head damage and I could find nothing but anecdotes (from good people, like Raby for example). 450 seemed to be the do not exceed value (although not a guaranteed death) 500 being the 'you have done something terrible', and 400-425 being the 'better watch that Mr" area. Pretty much don't cruise at anything 400 or above. I wonder about the absoluteness of CHT though. AFR and advance can all cause issues but I wonder if that is reflected 100% in CHT. For example, if you are not lean, and you are at 400 is everything fine and if you are lean and you are at 400 is everything not ok. Is it ok to be at 420 if you are rich? Or would you not have reached 420 if you were rich? Get what I mean? Should you feel confident that you are OK if the CHT says you are not in the danger zone or does that zone change depending on the tune. Seems to me that valve seats and stems are just gonna care about the temp.

And I really have to think about all those years of VW buses and 914s without CHTs, flogging the crap out of them, and still people feeling the engines were good and pretty well bullet proof.

It is indeed a very complex question with many variable parameters. If you really want to understand it better, I again suggest that you go to the world of aviation engines. The information developed by a group in Oklahoma is particularly insightful A good place to start is here

https://www.avweb.com/features_old/pelicans...8mixture-magic/

In the realm of aviation, they are generally not that concerned with burning valves or dropped valve seats. Detonation is the main concern You will have to wade through a lot of aviation oriented jargon, but it explains the relationship between mixture, CHT, EGT (directly relatable to mixture), timeing, power output etc. Remember - LOP refers to lean mixtures with EGT's below their peak value and ROP refers to rich mixtures with EGT's below peak on the rich side.

Problem is, with variable fuel injection (or carbs) constantly changing mixture, and without more instrumentation (AFR, EGT, CHT, MAP, knock sensors) you will never really know what is going on inside the combustion chamber. And in addition, an auto engine is a much more variable situation than an aviation engine, where variables change much more slowly and are pretty much constant for long periods.

So in our application you are basically left guessing and relying on the carbs or fuel injection to get it "right" and watching CHT as a rough, cumulative indicator.

The temp ranges you describe are pretty good and safe.

From 2014-16 I spent two years with a '75 914 that had a 2270 with a pair of DRLA40's, and a 4-channel digital MGL Avionics Stratomaster Velocity CHT designed for light aircraft. That car taught me alot about head temps. Along with the AFR gauge it had, it allowed me to really nail in the tuning of the DRLA's and the engine as a whole.

First off there are essentially two "zones". The crossover point between the two zones is between 60 and 65 mph. Below 60mph, air resistance and drag are not factors on head temps. Above 65mph, they heavily impact your head temps.

In the lower zone, the engine was coolest between ~2800rpm and ~4300rpm. You could get away with running a little below 2800 on a decline (maybe down to 2500), but if you started uphill then temps would slowly start to creep. The upper threshold was over 4k (maybe up to around 4500 depending on conditions), where sustained running would make the temps start to creep. Outside that window, head temp creep was directly analogous to how far outside the window you were: the further outside, the faster the creep.
In the "lower" zone I only ever used gears 1-4. 5th gear was verboten, and I'll explain that in a bit. My typical cruising head temps in this "lower" speed zone, usually cruising between 3k-4k were:
1) ~295
2) ~275
3) ~315
4) ~295
Those numbers are from memory and are not exact, but are in the ballpark. 1 and 4 were always very close if not the same, 2 was coldest, and 3 was hottest. There was usually 20-25 degrees between each step. I should also point out that with real time digital monitoring, the temps do not stay static. They are always moving around from second to second, but you get an idea of the average from the numbers they hover around.

However, once you cross the 60-65mph threshold things change. This is the speed where air resistance and drag begin to become real factors on the load the engine feels. My head temps were always about 30 degrees higher across the board going over 65mph on the interstate - with the top on. Take the top off and it would add another 15-20 degrees because the car's aerodynamic efficiency is now destroyed. In the "lower" zone below 60mph, it made no difference on head temps whether the top was on or off, but above 65mph it made a very real difference.
I never used 5th gear below 3000rpm, which on my car was 70mph, because head temps would slowly creep, even on flat grade running in 5th at say 2700 or 2800, which was like 67mph or so. And if I was going up hill in 5th at 3k, I'd usually drop a gear to 4th just to be safe. The 914, IMO, is best used as a 4 speed below 70mph. 5th should be saved for above 70 in all circumstances, unless you want to risk cooking your heads. Honestly I cringe every time I see a driving video on BAT where the seller is cruising around in 5th gear at like 60mph.
Also, above 65mph, the area of most cooling narrowed slightly to around 3000-4200. As with the "lower" zone, the further outside this window, the faster the temp creep, but with a sharper deviation.

But what was fascinating to me was how quickly temps could change. I'd be going down the interstate at 75mph with #3 cylinder at 355 degrees, and in the 10 seconds it takes to exit and sit at the stoplight at the end of the off-ramp, it would be back down to 315 degrees before I'd even come to a full stop.

Wow, that is some fantastic information. The zones bit is really interesting. I believe I really see the same thing. I had not at all thought about the top causing issues but of course it is obvious. Lots to think about. My temps on a 2056 are not that high for cruising but. 2270 is a different animal. The movements and variation you describe are dead on of course. I think I can get away with 5th just because the cooling system can keep up better with the lower power. I do notice the budget though, cruising at 60 in 5th is a steady state, I do not lose any temp in 5th in the lower zone. It seems like a slow build. So much to think about.

I wonder what the aero on my car contributes too... that front air dam probably doesn't help.

THIS.
I found this stunning as well. No matter where temps are, the change when going from cruise to full stop is amazingly quick.

Looking forward to getting some of my own data on this again but it'll be a few weeks before I can get back and get the 123 distributor in and squared away. I'd really like to get an AFM in as well - had one on a previous 951 - but not sure it's in the cards for the 914. we'll see how the 914 runs after the distributer change.