mark kibort

I, like most of you , know what the thermostat does, but didnt really understand the flow pattern, or how the thermosat allows for certain flow containment until the engine warms up. I was working on Scot's car to flush out the cooling system, to try a ceramic sealer used to fix cracked heads. K&W Nanotechnology sealer. It calls for a perfectly clean cooling system before treatment, then you run the car for 30min and drain the system, while those nasty blown headgaskets, cracked heads are sealed up by the "nano molecules" curing over night (12hours with a dry cooling system after draining the stuff out). it calls for the thermostat to be removed for the treatement, but we know that cant happen with a 928, without comprimised results.

anyway, while running the engine, i realized i needed to have flow to get the stuff in the engine. so, i took an old thermostat and "fixed it " open. I used a hobby splined gear for a minature electric motor that fit on the end of the part of the Tstat that fits in the housing when the t-stat is closed (cold) . anyway, it simulates it being open. however, i then started to look at how the plunger seals off the water coming from the heads. it didnt reach the opening totally, so i started to think. since we put the t-stat in boiling water and it moved X distance, that distance is not far enough to reach the opening to the heads coolant path. Hmmm. Is it suppost to be partially open?? some of the water that feeds the pump is hot from the heads and the other portion comes from the bottom of the radiator?? didnt sound right, so i found another gear that reallly opened up the t-stat and when i fitted it, it hit, by 1/4" too much, but its spring loaded and compressed as i tightened the t-stat housing around it.

I started the car, added the water and head fixer stuff and was amazed at what i saw as far as temps go. before that , i flushed the system and found the water pump is pumping about a garden hose hose flow strength as i kept a garden hose on a separate 928 hose leading to the new t-stat, and the water was exiting the bottom of radiator going to the street.
reved the engine, and water shot out at a high rate, at idle, i had to pinch the hose to match the water pump flow at idle. So, the pump is fine.

Scots car has been running hot anyway, since the rebuid. low and behold, the engine runs cooler, sure it took longer to heat up, but not that much longer. it spent quite a bit of time at below the mid line!! and with the fans on , the temp never went above mid way. keep in mind, this was a 90 degree F day too! certainly, this is no good for a street car, but im thinking for a race car, maybe this is the ticket to fight overheating. (manual pegging of the t-stat)

I noticed while working on my friends 85, that there is a nice seal that meets this opening for the head's coolant jacket. i suspect that with the 85s, maybe there is a better seal there. Or, maybe the t-stat gets hotter than when i was testing it on the stove in boiling water.

anyway, im seriously considering leaving the "pegged" t-stat in scots car as now the water is really circulating. and if the t-stat never bottomed out to the Head's cross over water jacket, where this would mix hot water with the radiator cool water, maybe we need a t-stat that moves farther inward to make the seal really happen.

any reason why this wouldnt work?

NOTE for those that dont understand the t-stat's job and design:
when cold , the t-stat blocks the inlet coolent hose from allowing any water to enter the engine to the pump. the pump takes water from the top of the T-stat housing being fed either hot water from the water in the heads, or when the t-stat is open and seals that port, takes water from the bottom of the radiator hose. the pump pushes the water to both sides of the block, and the water goes from the block upward to the rear of the heads, where then the water traverses the heads to the front of the engine where it meets the "head's coolant cross over" it then has an opening that faces the Tstat where if the t-stat is closed, it has a direct shot back to the water pump to circulate hot water and assist with warm up.

my only question, is wheter the water rises to water jacket ports along the length of the heads, or does the water flow around the cylinders to reach the rear port where the water goes up and then to the front of the engine side of the heads to then exit a port that connects to the top of the radiator, where it is cooled and then goes into the engine from the lower radiator hose.

Ok, thats 928 cooling 101.

comments?

any bets on wheter that head crack sealer will do anything??

Heads may be coming off soon. im knocking for scot, knock knock knock!

MK

SharkSkin

Mark, IMHO running any engine with the t-stat out(or in the 928, pegged closed) is a really bad idea. Drag racers make exceptions like this all the time, but then again they don't always use coolant. Sure, you may want to get all the cooling you can while on the track, but a properly operating t-stat will give you that when needed.

There seem to be some gaps either in your understanding of the way the 928 cooling system maintains temperature, or in your explanation of your understanding. It's not really an on-off switch, it mixes water from the bypass with cooler water from the radiator, as needed, to maintain a constant engine temp. Most of the time when running it isn't really open *or* closed. If it was not able to close off completely(when needed) that would account for Scot's overheating. But if you jam it closed then the engine will take too long to warm up, it will cool too much on pace laps, and the engine simply won't last as long because it will never really be at the proper operating temp.

I'd say just make it work as intended. If it wasn't closing all the way then you are missing a seal in the rear of the housing or you have a bad t-stat.

The system is a bit unusual. Rather than try to explain, Here is a page that explains the details. I've also posted blown-up diagrams.

mark kibort

Dave, thats exactly how i understand it. I think the difference is in the application, where i dont think there is too much of a difference on warm up. I think, at worst case, we are talking about 2x the warm up time, as there is probably more water in the block than in the radiator. I actually did the test yesterday. It warmed up a lot slower, but probaby at full temp at 15 min, vs normally at 5-8 min. (but i was also driving slowely too)

on the track, from the warm up lap to the cool down lap, temps are always normal, but the cooling system is fighting to keep it there due to the high heat of WOT. without a fan, you are toast unless you are really moving fast at a longer track.

anyway, no argument to the proper use of the t-stat. i was just thinking for the track, we could peg Scot's, if his isnt closing all the way. the early blocks dont have that seal at the t-stat hosing against the cylinder head cross over port. if not sealing, then there is always a chance of the HOT coolant slipping by and going directly to the pump, to short circuit the system. different degrees of this is nessesary for warm up and when you are not demanding full output. but at the track it probalby is not an issue. However, even my functioning t-stat cooling system drives as low as the lower white line on commutes to the track. we are thinking about putting a blanket in front of the ratdiator to keep the engine a little warmer, even though the t-stat is closed im sure.

The main issue for scot is the crack in his heads or head gasket. I think all of this is an excersise in frustration due to us not being able to regulate temps properly due to hot gasses heating the system and under race conditions, over pressurzing the system and pushing out most of the water , and then it over heats due to lack of water and pump impleller cavitation.

Im hoping the crack block treatment works, but if it doesnt,its OUT WITH THE ENGINE!! YET, AGAIN> we think that this was all caused by the cracked overflow tank that caused an overheat where the red light came on, and then scot didnt turn off the engine but continued to run it for 10 to 20 seconds until the light went out, and then came into the pits. this cracked the heads (i think) then, after we fixed the bottle, more overheating and loss of coolant due to the crack in the heads pressurizing the water jacket.

NOTE TO SELF : see red light, turn off engine!!!!!! (or approaching red area, turn off engine!)

MK





[QUOTE=SharkSkin]Mark, IMHO running any engine with the t-stat out(or in the 928, pegged closed) is a really bad idea. Drag racers make exceptions like this all the time, but then again they don't always use coolant. Sure, you may want to get all the cooling you can while on the track, but a properly operating t-stat will give you that when needed.

There seem to be some gaps either in your understanding of the way the 928 cooling system maintains temperature, or in your explanation of your understanding. It's not really an on-off switch, it mixes water from the bypass with cooler water from the radiator, as needed, to maintain a constant engine temp. Most of the time when running it isn't really open *or* closed. If it was not able to close off completely(when needed) that would account for Scot's overheating. But if you jam it closed then the engine will take too long to warm up, it will cool too much on pace laps, and the engine simply won't last as long because it will never really be at the proper operating temp.

I'd say just make it work as intended. If it wasn't closing all the way then you are missing a seal in the rear of the housing or you have a bad t-stat.

The system is a bit unusual. Rather than try to explain, Here is a page that explains the details. I've also posted blown-up diagrams.

ErnestSw

Mark,
I completely understand what you've said and I've also cranked a thermostat wide open to "cure" an overheating problem (that I didn't have). Teh results, in my case, were exactly the same operating temperatures when in Texas heat, but colder than normal temperatures if the ambient got below around 70F.
The problem with Scott's car is not caused by the thermostat. I don't remember which year cars added the seal behind the thermostat, but if Scott's car is supposed to have one and doesn't it could definitely cause his overheating problem.

Rufus Sanders

This is really interesting. Mark, I would bow to your racer's experience for sure. and I have heard of people actually plugging the opening to do what you suggest. (someone actually posted a photo of their t-stat housing with a rubber plug, the kind witht a little **** you turn to expand it to fit the hole, that had been put in by the PO, loosened over the years and finally fallen out and allowed the car to run hot again)

As for proper operating temp, I must agree that you really want the whole drivetrain at an even constant temp, since it's alum. and so prone to quick expansion & contraction. My car ran hot for years, not really hot, but at the top of the "normal " range. I finally put in a new (read "clean and unplugged by dead birds, leaves, bees, sticks, plastic bags acorns etc. etc....") radiator, flushed the system completely, and had the t-belt tension adjusted.

Now it runs so much smoother, you'd swear it was a new car. The temp is at about 82 - 84 C, and once warmed up, the needle never moves more than one width up or down. So, I'd say, let the t-stat do it's job, so you don't have to ever worry about it. Plus, they say the 9238 cooling system is actually very over-designed and maybe that's why we can have cars that "run hot" for years , but never really overheat or damage anything. - Ruf

mark kibort

you guys are probably right. good point about the rapid heating and cooling (more the cooling) of the coolant, when the engine is hot. probably more of an issue when the engine is hot, than when the engine is warming up at a similar rate. oil at 190f and then water goes from 180 to 70 and then you got all sorts of disimilar temps, expansion, etc.

Ernest, the problem is more than likely the cracked head or gasket. however, its so frustrating because it cant be seen as an issue, unless the car is raced! i think the wide open thermostat could give him a lap or two more time in the race, but either way he is in trouble and the engine has to come out. Unless the K&W nanotechnlogy works by a miracle, he will race at laguna and run 7 fast laps, and come in or cruise the remainder of the race and then it will have to come apart!
ahhh! and Ill be the one doing the work again!! 3 engine pulls for the guy!! heck, all that work, i could have fab'ed by own stroker by now!

MK

littleball_s4

Just a thought:

If you turn off the engine when the needle is in the red area, it will overheat massively and probably bend and crack everything.

IMO, the best thing to do is to lift completely and cruise to pit, better in fifth than in neutral, as if it runned out of gas, to mantain both air (because of car speed or fan) and coolant flowing. If it's too close to pits, I would even try to keep it going slowly in track if it's safe (free practice), or at least idle it with the bonnet open for a little while (as long as there is some coolant left. If it's not the case I think you've totalled the engine anyway).

The only case in which I would inmediately turn it off would be (apart from no oil pressure obviously) a power loss followed by slight overheating, which could mean pistons having problems or lubrication problems.

PorKen

Scot has a AFPR, correct? Dial in some more fuel - that'll cool it down.

My '81 runs cooler at cruise with the AFM bypass screwed CW all the way in (rich), versus the six turns CCW (14 turns max) which I run for DEQ tests.

mark kibort

we got scot pretty fat right now. in the 12:1 to 12.5:1 range

AFM bypass is only for part throttle as you know. the question is running at cuise and running leaner or richer. actually, either side of stoich can run cooler, but actually the coolest will be lean of stoich. just before the engine starts to miss, 17:1 or something, you can get best fuel economy and coolest EGT and CHTs. the afm bypass is such a fine tuning device. I dont think it will effect much during actual driving conditions, compared to the effects it has on emissions. the O2 sensor is fighting with the AFM (or working with ) to keep things at near 14.7:1 average. by playing with the adjustment screw, the mixture doesnt change all that much. you want to see REAL rich at cruise??? just disconnect the o2 sensor! it will run at near 13:1 all the time.

scots overheating isssues are really revolving around the crack in the head or head gasket. (probably the head)

Mk

mark kibort

actually, when the pump fails or cavetates (lack of water) the water is not moving. so, turning off the engine can be the only option. it wont get any hotter upon shut off, in fact, it starts to cool immeadiately. Only if you have water circulating,, can you continue to run at a lower power setting to cool things down due to overheating caused by timing, radiator flow issues or other things causing inadaquate cooling at full power settings. if you drain your water, and there is no flow, the engine will go in to thermal run away, and there is nothing you can do but shut it off. the longer scot or anyone runs the car,the hotter it will get until it blows!

now, scot does have a combination of issues. so, when he overheats, he has some flow, but not much, and at a lower power setting he can drop the temps to the upper range, but not in the red. (air fllow and some water circulation, as you say, does this).

when we think scot cracked his head, is when the redlight wouldnt go out even lower power settings and the light was on for too long. he had vaporized much of is coolant due to the cracked overflow tank during a race and lost most of the coolant.

Thanks.

MK

mark kibort

Well, we did the treatment and let it cure over 2 days, and then flushed the system, refilled and drove it. since we pegged the t-stat to do the flushing and treatment, we left it in like that so it took a while to heat up. when it did, i poped the cap and did a exaust gas check on the coolant reservoir. when i ran it, i did some WOT runs, and then came back. while i was testing it with the "crack block/head tester, i reved the engine a bit to see what the condition of the coolant was in . sure enough, there was no change of color of the test liquid of the kit. (when there is exhaust gases present, it turns from blue to yellow , like it did before the treatment. ) So, at first glance, it seems to of worked, but HOW!!!!???? If this works in a racing environment, ill be even more amazed!! we will see at the SCCA race at laguna seca on Oct 1st.

K&W Nanotechnlogy ceramic sealer for cracked heads and gaskets, etc.
Could be amazing stuff. we will see.

Ive been told the stuff works by attaching to the hot spots of the heads when cracked and being heated by the combustion chamber heat leaking into the cooling system. its suppost to be perminant too. hard to believe this stuff can work.

The first test detected exhaust gasses in the cooling system by turning the test liquid yellow. now it stays blue. do you believe in miracles!??

Mk

littleball_s4

I don't, but keep my fingers crossed!

One not too expensive thing you could do is buy a decent 0-20bar pressure sensor (shouldn't be more than 50$) from a equipment shop and put it in the coolant circuit. Add a gauge in the dash and you'll have some clues about what's going on:

1) if you never see more than 1.2 bars hot, the cap is not working right (or something else is leaking), coolant is boiling at 100ºC, causing overheating problems. Remember at 1.2 bars the circuit SEEMS presurized, but it IS not enough.
2) if you constantly see 2.0 bars, but jumps or vibrates to 20 when hot WOT conditions, it points to the feared crack.
3) if you normally see 2.0 but, in the middle of a race, it drops to 1.2, the cap was forced open and stayed like this beacause of whatever reason (I've seen it before), unleashing hell.

SharkSkin

I might, come Oct 2.... would be interesting to run that test again after the race. For now though, sounds hopeful!

mark kibort

I have a 1 bar sensor (sunx) very sensitve. , but i could get access to a liquid pressure sensor with higher PSI range. however the cap is rated at 1bar, thats 15psi. as soon as you get 1bar, the cap starts to open via its spring and air or coolant starts to leak out. however with our issues, the coolant is pressurized via the leak in the head and just starts to pus the actual coolant out of the cap (and over flow tank)

what we could do , before the pressure gets too high , is look for a pressure change when he his at WOT. if it goes up , we know its still broke! if it doesnt, but does later in the race, then we know it gave way later in the race.

Im thinking of putting a second over flow in the car to visually look at. when the gallon bottle gets near full, time to come in ! if he wants to risk it, he "yanks" the line off and going to the expansinon tank and continues on.

MK

littleball_s4

Oh, sorry, all pressures were absolute, not gauge. The cap is rated at 1 bar relative, so 2 absolute. Absolute sensors also allows you to see faulty caps that doesn't let air in, creating below one pressures when the engine cools.

Good idea the second tank.