Building a 10,000 rpm NA engine.

[bazza]

Originally Posted by Dessertrunner

Its not possiable to get 75% of pump flow unless the radiator is above 78C, sorry to be the bear of bad news but the amount the stat lets through is directly related to the amount coming round the side pipes and its temp.

Agree with everthing else you said.
Tony

Yes - but you're assuming the thermostat snaps shut - but we're not sure how long it takes to transition and what the flow rate effects are.

It could be fully open with 75% through the "radiator" for 2-3 seconds and then taper down to 0% if it was a once off burst of engine load / hot head temps.

Or if the engine is loaded up (doing laps) it might be going between 50% to 75% flow for example - like a sine wave.

[SVXRide]

Originally Posted by bazza

Yes - but you're assuming the thermostat snaps shut - but we're not sure how long it takes to transition and what the flow rate effects are.

It could be fully open with 75% through the "radiator" for 2-3 seconds and then taper down to 0% if it was a once off burst of engine load / hot head temps.

Or if the engine is loaded up (doing laps) it might be going between 50% to 75% flow for example - like a sine wave.

Car thermostats are definitely not fine Swiss watches

Should be easy enough to set up some pots on a stove with water at different temps to see what the response time is.

Bill

[oab_au]

Originally Posted by Dessertrunner

Sorry to everyone with out access to this mag Race Engine Tech,

Harvey on page 26 top box of the August issue it says the rules limit the engine to 7,500 rpm. Also as I understand it we are talking about a turbo engine not NA so I am not sure why you think it best to run the engine at 7,500 rpm.
In the Jun/July issue it has a story about the development of a motor for the V8 super cars (Australian version of NASCAR). What's interesting is that the torque curve page 62 is flat from 5,800 to 7,500 rpm which is impressive and they have again rev limited the engine to 7,500 rpm due to the rules.

In every case the engine have been pulled back by rules not by the engine developer. I would like to understand why you want to us to pull the revs back or develop maxium power down lower. I want the motor revs to go higher if I could get it to run at 12,000 rpm I would do it. Untill the race rules pulled these guys back they were going as hard as they could to move the revs up to get more power with a NA engine.

I need to understand why you think its better to pull revs back, when I do the maths it just doesn't make sense a 3.3l engine will be lucky to get more torque then 350 nm, (please jump in if you think we can get a higher torque number)
If this torque is developed at 7,000rpm we will get 256kw
if the same torque is developed at 10,000rpm we get 366kw
if I develop the same torque at 12,000rpm we get 439kw.

I am happy for you to show me why I am on the wrong parth with higher revs, we need to be going higher not lower. This point has been buging me for a while so I would like to get some sort of understanding about were you thinking is. I accept that we will need to try to get some sort of flat torque curver to make it all work but that is a seperate issue. We don't need to start the project assuming that a flat torque curver is impossiable.

Please don't take it that I am having a go at you I am only trying to understand were you head is at.

Tony

Tony there are really two reasons that the engine should be developed, to run at the lower rpm range of 7500 max.
First, is the mechanical problems in the rods and pistons surviving the high inertia loadings that a piston speed of 25mts per sec. will have at 10000 rpm. The valve gear needs to be very light to try to keep the valves from floating into the pistons and each other at that speed.

Secondly, you are treating the torque curve and the Max rpm like two are separate issues, they are the same thing, the HP is the result of where, and how much torque is produced.

To develop the HP at 10000 rpm the torque of 350nm would need to be developed at 9000 rpm. At those rpms 350nm is not attainable with this engine. As the engine speed goes up, the time to fill the cylinder decreases, we can use as much duration and sonic pressure as we can, to do it, but the torque curve will become peaky and have a very narrow curve, that will only be usable at 8500 to 9500.
Below these rpms the torque wouldn't pull the skin off a rice pudding. We would need a very low diff gearing of about 6:1 and a multi close ratio gear box. It would not be usable for the type of running that the car is to do.

You are just starting at a number, 10,000 rpm, then wanting the engine to do it. I believe it is better to develop the engine to have a strong, wide torque curve, that covers the usable range, given the road speed and the gearing, that the car will run.
High HP is only good for the drag strip, where the throttle is WFO all the time. On the track you need torque right across the rev range that the car runs.

As far as getting the water pump to work at 10000 rpm. It won't do it, it will just spin in its own vacuum. The pump impeller needs to be designed to run at that speed, and then it won't work well at lower speeds.
For that speed I would use an electric pump, that can pump the water, without running out of its speed range. Read Issue September/October, Water pumps.
The detonation problem that the supercharged cars had, was not due to heat, but to the Igniton timing that Mike programed the ECU to run.

Harvey.

[Dessertrunner]

The point at which a stat opens and closes doesn't change its 78c on ours as you said Bazza. The stat will start opening (I accept its a slow process) until the temp of the water around the stat see coolant below 78c . When it sees that it stops opening and start closing.

How do I know this well, I alway have a hole drilled in my stat because years ago I was in the snow and I watched the engine temp moving 20c. THe engine would heat up then the stat would open sending a fix of cold water which came from radiator and may have been at 10C, agree that this took a couple of minute but then the cold water hit the stat and it would close again. My wife and I watched this happen at a cycle time of about 5 minutes. The interesting thing is this did not show of the dash temp guage.

I need to apolize because I am not explaining myself properly.
What normally happens is if the radiator is below 78C then the thermostate will adjust the amount its open so that the mix of coolant it sees is hotter then 78C. It will never fully open if the radiator is lower temp then 78c. The further it is below 78C the less it will open which intern means less flow till it gets to a point like in the snow when it will cycle.

My point is the better the radiator / colder the radiator the bigger the problem become from lack of coolant flow. THe cooling ability of the system is litrs of flow times the differential temp.
Tony

[Dessertrunner]

Harvey I read what you say and see what you say but in the real world of race engine I don't see it backed up by actions. Nascar engine are limited at 10k is I rememeber correctly and even they want the power out there.

Water pump we are working on but pretty sure the electirc has got a chance of cooling the engine. There flow is rated on zero head pressure. Bazza and I have agreed that we may have to move the existing water pump to be able to control speed of the pump and also take pressure off the timing belt.

Tony

[1986nate]

Originally Posted by Dessertrunner

Harvey I read what you say and see what you say but in the real world of race engine I don't see it backed up by actions. Nascar engine are limited at 10k is I rememeber correctly and even they want the power out there.

Water pump we are working on but pretty sure the electirc has got a chance of cooling the engine. There flow is rated on zero head pressure. Bazza and I have agreed that we may have to move the existing water pump to be able to control speed of the pump and also take pressure off the timing belt.

Tony

Nascar is very different than road racing or running through the desert and also very similar to drag racing in essence. They are running WOT nearly a whole race, coasting through the corners. Even on the smaller tracks that are tighter and speeds only reach 130-140mph, they have different engine builds and different gearing (They stay in 4th gear all the way around the track) so that they are never, or very rarely, falling below 7000 RPM on any given track. They also use the engines for one race weekend(qualifying, practice, race) before it is completely rebuilt.

[oab_au]

Originally Posted by Dessertrunner

Harvey I read what you say and see what you say but in the real world of race engine I don't see it backed up by actions. Nascar engine are limited at 10k is I rememeber correctly and even they want the power out there.

As far as getting the water pump to work at 10000 rpm. It won't do it, it will just spin in its own vacuum. The pump impeller needs to be designed to run at that speed, and then it won't work well at lower speeds.
For that speed I would use an electric pump, that can pump the water, without running out of its speed range.

But Nascar cars run it a constant speed, they don't have to break hard for a slow corner then accelerate out of it. A rally car has to. What rpm range do rally cars run? Matt reckons 7500 would be the most he would run

Water pump we are working on but pretty sure the electric has got a chance of cooling the engine. There flow is rated on zero head pressure. Bazza and I have agreed that we may have to move the existing water pump to be able to control speed of the pump and also take pressure off the timing belt.

No matter where you move it to, it will have to operate from 2000 rpm to 10000 rpm.
Tony

Read the article in the September/October issue.

Harvey.

[bazza]

Originally Posted by oab_au

Tony there are really two reasons that the engine should be developed, to run at the lower rpm range of 7500 max.

First, is the mechanical problems in the rods and pistons surviving the high inertia loadings that a piston speed of 25mts per sec. will have at 10000 rpm. The valve gear needs to be very light to try to keep the valves from floating into the pistons and each other at that speed.

These issues can be easily solved - Pauter rods are known for being unbreakable + the valve gear which we got ordered is far lighter than the original hydraulics and other STI solid buckets for example.

Secondly, you are treating the torque curve and the Max rpm like two are separate issues, they are the same thing, the HP is the result of where, and how much torque is produced.

To develop the HP at 10000 rpm the torque of 350nm would need to be developed at 9000 rpm. At those rpms 350nm is not attainable with this engine. As the engine speed goes up, the time to fill the cylinder decreases, we can use as much duration and sonic pressure as we can, to do it, but the torque curve will become peaky and have a very narrow curve, that will only be usable at 8500 to 9500.
Below these rpms the torque wouldn't pull the skin off a rice pudding. We would need a very low diff gearing of about 6:1 and a multi close ratio gear box. It would not be usable for the type of running that the car is to do.

You are just starting at a number, 10,000 rpm, then wanting the engine to do it. I believe it is better to develop the engine to have a strong, wide torque curve, that covers the usable range, given the road speed and the gearing, that the car will run.
High HP is only good for the drag strip, where the throttle is WFO all the time. On the track you need torque right across the rev range that the car runs.

350 Nm @ 9500 rpm is only 340 odd flywheel kw if I'm not mistaken - that's only 260 kw atw which I think is very achievable with that rpm... Jack's EG33 made close to that already and it had a nice fat torque curve.

As far as getting the water pump to work at 10000 rpm. It won't do it, it will just spin in its own vacuum. The pump impeller needs to be designed to run at that speed, and then it won't work well at lower speeds.
For that speed I would use an electric pump, that can pump the water, without running out of its speed range. Read Issue September/October, Water pumps.
The detonation problem that the supercharged cars had, was not due to heat, but to the Igniton timing that Mike programed the ECU to run.

Harvey.

You'd need an aviation spec electric water pump - even 2 of the biggest Davies Craig pumps drawing massive amperage will just get close to the stock pump. Tony and I have come up with the idea of the following. Gut the stock pump so the idler still turns. Feed that gutted pump from a modified stock pump mounted in the aircon location driven off the aircon pulley. May need to modify gears but the stock pump is very good so we just need to slow it down which is rather easy given our idea.

Detonation and heat tend to be joined at the hip. A hot engine will detonate far easier than a cooler engine and a good aftermarket ecu has inlet temps and engine temps accounted for - mine for example goes into limp mode above 100'C and pulls timing anywhere outside of the acceptable zone.

[Dessertrunner]

The dry sump will be installed on Bazza car and then we will move onto developing a new top pipe for the cooling. Is important that one stage is completed before we move on. After we develop the top cooling pipe with twin stats we will then start the test bench work on oil flow and coolant pump supply as we have discussed before. What I don't think we mentioned was that we built 2 dry sumps while we were going.

When we get the test bench going it will answer some of the question. After we are happy with the oil and cooling system we will fire up the dyno and run some torque curves.
Tony

[bazza]

Originally Posted by Dessertrunner

The dry sump will be installed on Bazza car and then we will move onto developing a new top pipe for the cooling. Is important that one stage is completed before we move on. After we develop the top cooling pipe with twin stats we will then start the test bench work on oil flow and coolant pump supply as we have discussed before. What I don't think we mentioned was that we built 2 dry sumps while we were going.

When we get the test bench going it will answer some of the question. After we are happy with the oil and cooling system we will fire up the dyno and run some torque curves.
Tony

With the top cooling pipe - 29 mm exits from each head. This is 660 mm^2 of area x 2 = 1212 mm^2. Previous issue is that this is then trying to squeeze out the cross over pipe exit (next to PS pump) which is 707 mm^2... just a nasty little setup really.

So using my 46 mm radiator inlet up top with a merge just before the inlet - that's 29mm x 2 (1212 mm ^2) going into 46 mm (1662 mm^2 area)... certainly not rocket science to see that this basic mod will increase flowrate dramatically and reduce the pressure here.

[oab_au]

Originally Posted by bazza

These issues can be easily solved - Pauter rods are known for being unbreakable + the valve gear which we got ordered is far lighter than the original hydraulics and other STI solid buckets for example.

Well 10,000 should be no problem for you.

350 Nm @ 9500 rpm is only 340 odd flywheel kw if I'm not mistaken - that's only 260 kw atw which I think is very achievable with that rpm... Jack's EG33 made close to that already and it had a nice fat torque curve

Well looking at that torque curve, even if the numbers are a bit inflated by the dyno, it still makes peak torque at only 5750 rpm.

You'd need an aviation spec electric water pump - even 2 of the biggest Davies Craig pumps drawing massive amperage will just get close to the stock pump. Tony and I have come up with the idea of the following. Gut the stock pump so the idler still turns. Feed that gutted pump from a modified stock pump mounted in the aircon location driven off the aircon pulley. May need to modify gears but the stock pump is very good so we just need to slow it down which is rather easy given our idea.

There is no dtifference in power needed to drive either mechanical or electrical pumps for the same output. The mechanical pump on a small block Chevy takes 2kw at 4000, yet 10kw at 6000 rpm. How high are you going?

Detonation and heat tend to be joined at the hip. A hot engine will detonate far easier than a cooler engine and a good aftermarket ecu has inlet temps and engine temps accounted for - mine for example goes into limp mode above 100'C and pulls timing anywhere outside of the acceptable zone.

Well we are talking about engines that were built and run by members here, not your one. These had nothing to do with engine heat.

Harvey.

[Dessertrunner]

Wow what a debate, its all good info.
Harvey I don't want to debate back and forward for ever I just want to say that the only way I am going to get this engine to run at 10k is if I beleive its posssiable.

Even if its just a wild dream;-

I want to fail from trying,
not instead,
Not try because I might fail.

We have a saying at my work,

The possiable is easy,
the impossiable just takes a little longer.

This engine is just part of that rule.

Were going to 10,000rpm plane and simply, end of discussion, its going to happen I don't care what it takes or how long it takes me it will happen.

Have a great day all.
Tony

[bazza]

Originally Posted by oab_au

Well looking at that torque curve, even if the numbers are a bit inflated by the dyno, it still makes peak torque at only 5750 rpm.

You should do some more digging and research into Jack's setup. Do you see anybody in motorsport actually advertising their race power.... GOSH what a thought.

Now, moving to the torque, even if we have a tapering torque curve it's not an issue, so even if torque peaks at 6000-7000 rpm as long as the power keeps climbing and the torque isn't dropping off to quickly it's still going to be ridiculously quick.

This is an example of one of my old setups from about 5 years ago - the torque curve tapers however this car was ridiculous. Low 11s @ the drags, 2nd fastest Liberty around Winton (still to date) and from the apex out it wanted to light up the wheels, so much power everywhere. Drove it as a daily driver for 100,000 kms. Nobody with any sort of clue can say a tapering torque curve doesn't work - it's just utter rubbish. You just don't want it looking like a mound where it comes up peaks and falls straight back down.

There is no dtifference in power needed to drive either mechanical or electrical pumps for the same output. The mechanical pump on a small block Chevy takes 2kw at 4000, yet 10kw at 6000 rpm. How high are you going?

When we transfer from mechanical energy to electrical energy back to mechanical energy we lose energy each step. Remove a step - remove the losses - even if it's a few kw, it's everything for Tony's engine as we want everything out of it. Plus no point stressing the alternator if we don't have to.

Further find me a bullet proof electric water pump under $1000 which can do the job of the stock pump, 200LPM etc etc. Davies Craig only go to 115 (which I have one of them to run my rear water to air intercooler and I doubt it would even get to it's rating).

Further again - adding an electric pump adds another single point of failure, something avoided when you build a car to go racing with.

[Dessertrunner]

Here is the computer simulator I promised for the cooling system on a SVX, its excel so play with it.
- Yellow cells are entry.
- When the cell turns red it means you are outside of possiable range.
- It assumes the thermostate is in place.
- The name of the game is to get the highest number in the cell that is headed "Amount of Heat removed per minute in Kcal" the bigger the number the better your engine will cool.
- Thermostate opening temp is fixed at 78 may change it so you can install a different opeing temp but lets see how much interest is in it.

Have a great day all.

Tony


PS had to fix file

[Dessertrunner]

So a few of you have down loaded the cooling simulator, have I caused more confusion or does everyone think its totally wrong.
It would be good to get some feed back or discussion going so we can all agree if I am on the right track.
Tony