Low Temperature Thermostat (LTT), Bore Scoring (short version) and 3Rd radiators.

The problems with the M96/7 engines and their influence by the above features are best or benefit from being discussed together because they all impact on each other.


THE THERMOSTAT


An engine thermostat, has two separate but interrelated functions:
1. To bring the engine up to optimum operating temperature as quickly as possible; and
2. To maintain the engine at optimum operating temperature thereafter.

Introduction

To start with it is essential to accept that our type of Porsche engines create very different temperatures depending on largely how hard they are being driven.

From around 30 BHP in built up areas to over 300 BHP on a track or race day the job of cooling the engine varies enormously.

When an engine burns fuel, heat is produced. This heat increases the pressure of the resulting gas mixture, the remains of the intake air and the burnt fuel vapours, which forces the piston down and turns the crankshaft. But not all the heat produced by burning the fuel is turned into useful work; some of it remains in the gas and goes down the exhaust pipe, and some passes into the walls of the combustion chamber and cylinder and is removed by the engine cooling system.

So how much heat is involved? Internal combustion engines are quite efficient at turning heat into useful work at the crankshaft; in the most efficient high compression diesel engines the "thermal efficiency" (the amount of energy in the fuel that is turned into useful work) can approach 50% under ideal conditions. Petrol engines are not as efficient as diesel engines because of their lower compression ratios, and most engines certainly do not work under ideal conditions most of the time. So let’s talk in round numbers: in a typical car engine, very roughly about one third (33%) of the energy in the fuel is turned into useful work to make the car go, one third of the heat goes down the exhaust pipe in the hot exhaust gas, and the final one third of the heat goes into the cooling system.

Using these proportions, we can see that a car engine, let's say a reasonably powerful one of 300 bhp (brake horsepower) when it is being driven flat out is actually making 900 hp of heat when it is producing its maximum power. One horsepower is approximately 750 watts so this is about 650 kw (kilowatts), or as much as 200 electric kettles. One third of this 650 kw heat has to be removed by the cooling system.

Of course an engine is not producing its maximum power all the time. When it is ticking over doing no work it is producing very little heat, and the demands on the cooling system are much lower. This is evident when you are sitting in a traffic jam, even on a hot day. In a typical modern car with an electric radiator fan, the temperature of the engine will gradually rise until the fan kicks in to cool it down. Once the coolant has been cooled a few degrees the fan shuts off and the coolant can then absorb the heat output of the engine for a few minutes until it has heated up and the fan needs to kick in again to cool it down.

So the engine cooling system needs to able to remove its part of the heat being produced by the engine, either a lot or a little, and anything in between, whilst keeping the engine temperature stable at its optimum operating temperature.

Using radiators to cool is a convenient idea but then their cooling capacity and ability is influenced by the air temperature (which varies hugely) and the air flow through them (which varies with both toad speed and what you might be following behind).

So with air flow from say 20 to 160 mph, temperatures from zero to 30 degrees C and engine power performance varying by 10 times from cruising at 30 mph and racing at well over 140 mph – it is impossible for one system to cater for all those variables perfectly.
If the system is then fitted with a coolant pump that varies speed with the engine – that introduces yet another variable.

If controlling all these variables is left just to a thermostat (and perhaps radiator fans in extreme conditions) that is never going to be good enough to cope with the multiple thermal variables described above and is always a compromise based on cost and the more powerful the engine the bigger the variables and the more difficult accurate control in all circumstances becomes.

To investigate how the M96/7 range cool almost 20 years ago we fitted temperature sensors inside the engines and radiator blinds and conducted tests in varying weather and driving conditions.

We found in some conditions (cold day - moderate speeds) 1/3rd of 1 radiator was sufficient whereas on the race track on a hot day and following close behind another competitor an additional 3rd radiator was required – a range of about 10 times the radiator area open to air flow – to manage all the variables - and all left mostly to one thermostat to control.

To do the job better would require variable digitally controlled radiator blinds (or air flow dams) and variable electric coolant pumps (or control valves) all much too expensive for commercial use.

What made things even more complicated with our engines was the thermostat location.

Traditional technical information about the way thermostats control engine temperatures and the typical settings are usually written on the assumption that the thermostat is located at the exit from the engine. This is a really important distinction rarely mentioned because before 2000 most engines did have their thermostats located there.

In all engines there is a temperature rise as the coolant passes through the engines (matched by the temperature fall as it passes through the radiator) and so where a thermostat is located dictates the inlet temperature and if it is moved it changes it.

If (for example) an 83 deg thermostat (std Porsche 006/7) was placed at the entry to the engine then all temperatures inside the engine will be higher than that as the coolant passes through it and collects heat and so the exit temperature will be higher.

If the same 83 deg thermostat was placed at the exit from the engine them all the temperatures inside will be lower than that because on exit the coolant already has had the heat the coolant has absorbed on its passage through the engine removed added to the entry temperature and is now exiting hotter.

It is therefore essential when discussing thermostats, what they do and how - to state their position in the engine first (which few reports ever do).

There is a general agreement that Internal combustion engines operate most efficiently at relatively high temperatures, typically above 80°C - 85°C (176°F - 185°F) to a maximum of 105 (220). Wear on the moving parts is reduced and thermal efficiency is increased by operating at around this temperature.

Lower engine temperatures result in inefficient combustion which causes increased fuel consumption, and increased wear with consequent reduced engine life.

However, if the engine temperature gets too high, boiling of the coolant leads to local steam pockets forming which severely reduce heat transfer in the affected area (usually the cylinder head) resulting in premature combustion of the fuel air mixture, also known as detonation or knock, and can ultimately damage engine components (the cylinder head, valves and pistons).

To remove heat from the engine block and head, coolant is circulated in passageways cast into those components. Some of the coolant is recirculated around the engine, and some is diverted through the radiators to be cooled. The proportion of coolant recirculated round the engine versus the proportion sent to the radiators and cooled is determined by the degree of opening of the thermostat.

To ensure that an engine is always operating at its optimum temperature, the thermostat modulates its opening to control the flow of coolant, and consequently the flow of heat, from the engine to the radiators. Coolant is cooled in the radiators and returned to mix with the coolant circulating around the engine to maintain a constant mixed temperature.

If the engine is producing little heat, for instance if it is idling, then a trickle of coolant through the radiators is sufficient to remove this heat and keep the engine at a constant temperature. If the engine is working hard, then more heat is being produced and more coolant must be circulated through the radiators to prevent overheating.

External temperature and vehicle speed, which change the ability of the radiator to reject heat, also affect the rate at which coolant must be circulated through the radiator, because they affect the temperature of the coolant returned to the engine from the radiators to mix with the coolant circulating round the engine.

To ensure that the engine reaches optimum operating temperature as quickly as possible, the thermostat restricts the flow of water from the engine to the radiator to virtually zero (a small flow is required so that the thermostat experiences changes to the water temperature as the engine warms up) until the engine reaches optimum temperature. The thermostat then gradually opens up to allow sufficient coolant to flow through the radiator to remove the heat being produced by the engine and prevent the temperature rising higher. If the engine is warming up whilst idling, and consequently generating only a small amount of heat, the thermostat will only need to open a little to remove the heat being generated.

With the engine at optimum temperature, the thermostat controls the flow of coolant to the radiators so that the engine is maintained at optimum operating temperature, even as the power output, and therefore heat output, of the engine changes with varying load and ambient conditions.

Under peak load conditions, such as labouring slowly up a steep hill on full throttle whilst heavily laden on a hot day, the thermostat will be approaching fully open because the engine is producing maximum power, the velocity of air flow across the radiator is low, and the temperature differential between the radiators and the cooling air will be low. (The velocity of air flow across the radiator and the temperature difference between the radiator and the cooling air have a major effect on its ability to dissipate heat.) Note that even with the engine operating at full power, the thermostat should not be fully open: there should always be a reserve margin of cooling capacity on the precautionary principle.

Conversely, when cruising fast downhill on a motorway on a cold night on a light throttle, the thermostat will be nearly closed, because the engine is producing little power and the radiator is able to dissipate much more heat than the engine is producing. Allowing too much flow of coolant to the radiators would result in the engine being over cooled and operating at lower than optimum temperature. A side effect of this would be that the passenger compartment heater would not be able to put out enough heat to keep the passengers warm.

The thermostat is therefore constantly modulating, that is, it is moving throughout its range in response to the temperature of the coolant flowing past it, increasing or decreasing flow of engine coolant to the radiator in response to changes in the temperature of the coolant due to changes in power output to respond to vehicle operating load, vehicle speed, and external temperature, always keeping the engine at its optimum operating temperature.

A typical thermostat has a cylinder containing a heat sensitive wax and a piston which passes through the cylinder wall, to which is attached the operating disc of a valve and a return spring. Expansion of the wax as it is heated pushes the piston out of the cylinder, moving the disc of the disc valve. Contraction of the wax on cooling allows the piston to be pushed back into the cylinder by the return spring. At temperatures below the engine operating temperature range the wax is solid and the thermostat does not respond to changes in temperature. After the engine has been started and the coolant heats up, the wax liquifies when the temperature reaches the bottom of the operating temperature range. When the wax liquifies, the thermostat is at the point at which the piston begins to move the disc valve and divert some coolant flow to the radiators. As the engine warms up further, a steady flow of some of the coolant to the radiators removes surplus heat from the engine.

The wax expands or contracts in proportion to the temperature change, pushing the piston out of the cylinder or drawing it in with the aid of the return spring. The disc valve acts as a proportional control valve, controlling the proportions of the coolant that either recirculates directly to the engine, or is sent to the radiator to be cooled and then mixed with the recirculated water.

The thermostat is designed so that it will go from fully closed to fully open over a small temperature range. The temperature rating of the thermostat e.g. 71, 83, 88, 92 etc. is the nominal temperature in degrees centigrade at which the thermostat valve will start to open once the engine has warmed up. If the temperature of the coolant increases further, the valve will open further until fully open. The fully open temperature is normally 12-15 degrees above the opening temperature.

To test a thermostat, it is common practice to put it into a pan or kettle of water and bring this boil, observing that the disc valve goes from open to closed. However, testing like this can lead to misunderstanding about how the thermostat operates. The thermostat is designed to keep the temperature of the engine within a narrow range, and it does this by going from fully closed to fully open in over a temperature range of a few degrees.

Until the initial opening temperature of somewhere over 80 degrees is reached nothing happens, but once the opening temperature is reached the temperature of the water can rise so quickly through the thermostat's operating range that the proportional opening of the disc valve is not observed. This is why people mistakenly think that thermostats snap from closed to fully open in one step. To actually observe the proportional operation of a thermostat, the temperature of the coolant in which it is being tested should be raised very slowly over the operating range.

When testing higher temperature thermostats it should be noted that an 83 degree thermostat will not be fully open until 95-98 degrees, likewise a 71 degree thermostat will not be fully open until 73-76 degrees. The valve will not open fully when immersed in plain boiling water alone because the boiling point of water is 100 degrees centigrade at sea level. To test higher temperature thermostats they need to be heated in a water/ anti-freeze mixture or cooking oil, which will allow the temperature of the coolant to be raised above 100 degrees.

The boiling point of coolant increases with a pressure rise (which the closed system has) so is actually higher than it would be in open air.
Fitting the thermostat before the flow into the engine (or after it) makes little difference to the average temperature in the engine but does alter which part of the engine is running at the stated temperature rating or above it or below it. The engine on average runs hotter with the thermostat fitted at the entry of the coolant compared to the same thermostat fitted at the exit.

Despite carrying out extensive tests - with all the different ambient and driving conditions it is impossible to provide perfect averages but generally we found that the temperature rise driving on public roads driven slowly was around 2-5 degrees and driven more spiritedly was around 10-15 degrees C. This is where the role of the thermostat and coolant has to be considered for what it is trying to achieve in terms of oil performance.


BORE SCORING INFLUENCES


More importantly we found the temperature is different at the thrust face of bank 1 (of 3 cylinders) to the other bank 2 because the coolant enters at the bottom of both cylinders and increases in temperature as it passes upwards – but the thrust face on bank 1 is at the bottom and on the top of bank 2 meaning that where the piston is pushing hard against the cylinder wall (to drive the crankshaft round and the car along) it is hotter on bank 2 thrust face than bank 1 by around 5 degrees C with the standard thermostat.

It is very important at this point to understand that oil as a lubricant is supposed to provide a barrier between two metals that are sliding or rotating against each other to produce work. If the metals touched they would overheat, expand and the damage caused would result in crankshaft bearing failures and cylinder scoring or seizing.

Oil is provided under pressure to highly stressed areas (like the crankshaft bearings) under high loads to keep the metals apart but is only splashed onto the cylinder walls as the piston rises and remains there as the piston falls, dragged down by the rings forming a minor dynamic pressure rise.

Whereas the pump is always providing oil under pressure that rises in volume with the revs of the engine – the piston only has the clearance between it and the cylinder wall as the piston descends and all the time the forces driving the car are trying to squeeze the resulting oil film out.

Now pistons may have only 1.5 to 2.5 thou clearance (0.04 to 0.06mm) and that is equally shared each side (so only half that oil film can be present each side).

But the piston is not round or parallel but is oval, barrel shaped and tapered so most of the piston has more clearance and all the loads and the area close to the cylinder wall is a much smaller patch with spaces all round it for the oil to squeeze into easily.

At slow speeds (say 1,200 rpm) the piston takes 6 times longer to descend than at peak revs (7,200 rpm) so there is 6 times as long to squeeze out the oil film from where it was just splashed there when the piston was at TDC than where it is continually replaced with oil under pressure from the pump elsewhere in the engine.

So under high loads and low speeds the oil film between the piston and the cylinder wall is under threat of being squeezed thinner which there is 6 * less time to do at 7,200 rpm.

The viscosity of oil is measured by how long it takes to leak through a hole – the higher the viscosity, the thicker it remains at elevated temperatures and the longer it takes to drip through the hole (or squeeze out of a gap under load).

It is an established fact that ban 2 (the bank under a higher coolant temperature on the thrust face than bank 1) scores bores long before bank one follows – why is this?

It is also an observed statistic that on average Tiptronic examples score bores sooner than manual gearbox cars – why is this?


PUTTING IT ALL TOGETHER


If we accept the findings that the thrust face of bank 2 runs hotter than the thrust face of bank 1 AND that this makes the oil thinner on the face that scores on bank 2 AND that Tiptronics fail sooner (where higher torque is generally used setting off in 2nd gear than manual cars setting off in 1st) it is impossible not to be able to recognise that there is a connection between the oil temperature and how long the piston to bore surface finish survives.

It is also impossible not to recognise that where the coolant (and therefore oil) temperature was lower the cylinders and pistons lasted longer before scoring.

This then raises another mystery – why is the oil temperature (and therefore the oil film thickness and viscosity) related to bore scoring?


DETAILED ANSWERS IN OUR SEPARATE REPORT ON BORE SCORING


Briefly with age the aluminium in the cylinder bore wears and exposes the small hard silicon pieces that were spread out in the aluminium to provide a hard cylinder surface with oil pockets between the silicon that will expand and contract with the piston and dissipate more heat than with iron liners.

When the alloy wears away (by hydraulic erosion) the silicon particles are more exposed (stick out more) so they can penetrate the oil film if it is too hot and too thin especially at lower revs and higher thrust loads that squeeze the piston against the cylinder wall, squeezing out the oil quicker and with more time to do so - reducing the gap between them even further.

This didn’t matter when the early pre 2001? pistons had a very hard thin iron coating that resisted the scratching that results but the forced change to plastic coatings does not resist wear for as long and once it is worn off, flaked off, peeled off or ripped off (see report on bore scoring for pictures) the wear accelerates.

Specialists disagree with what happens next.

Some argue that the wear continues and somehow mysteriously causes score lines up and down the cylinder walls and pistons.

We believe that small particles of silicon eventually break free from the cylinder wall and are temporarily trapped between them before washing away (especially as they are sharp, pointed and can be bigger than the cylinder clearance (ref manufacturers own specifications and our report on bore scoring) and are pushed up and down the cylinder bore where they scrape out more material (now aluminium plus silicon particles) that spread wider and deeper until they create a big enough groove (or grooves) to do no more damage and eventually get washed away with the oil to the filter housing..


CONCLUSIONS


Because we measured a difference of between 2 and 5 degrees C in coolant (and therefore oil) temperatures between bank 1 and 2 thrust faces (bank 2 being hotter), a shorter survival mileage of the piston and cylinder condition between them and shorter in higher loaded Tiptronics (at low speeds where there is more time for the oil film to squeeze out thinner) we conclude that lowering the coolant (and therefore oil) temperature by 2 to 5 degrees must extend the life of the cylinder bores and pistons and by lowering it by 10 to 12 degrees – an even longer life could be anticipated before scoring (and that is supported by the mileages those who fitted our LTT are achieving).

When we fitted a LTT to test the same temperatures all over the engines we also found it reduced the difference between bank 1 and 2 to nearer 1-2 degrees - actually lowering the fragile bank 2 thrust face temperatures even lower and keeping the oil thicker – (but we are not sure why but it could be connected to the overall flow characteristics or that the oil cooler is also located on bank 2 feeding more hot oil to the galleries).

Our overall conclusion is that fitting a LTT extends life to bore scoring (but accepts it cannot eliminate it – only extend it – and we cannot yet state by how much because the distribution of the silicon particles inside the aluminium matrix varies between engines - although the cars we look after that have been fitted with a LTT do seem to have significantly longer mileages before bore scoring).


FITTING A 3RD RADIATOR (MORE DETAILS ALSO AVAILABLE IN A SEPARATE REPORT)


Our conclusion was simple - the third radiator in a manual car is only beneficial if the weather and conditions mean the existing 2 radiators cannot keep the temperature below the thermostat setting (particularly a problem racing following close on the fender/bumper of the car in front for a few laps).

Under normal driving conditions in a manual car - two radiators are usually sufficient but of course the 3rd radiator in a Tiptronic is piped to the Tiptronic gearbox where it removes heat from that to keep it cooler – so it cools an additional heat source and therefore does not overcool the engine.

Fitted to a manual used as an extra cooling source just for the engine - it provides more cooling than is needed except in extreme conditions when it is beneficial but then at most other times it overcools the coolant.
The problem then is the location of the thermostat.

By the thermostat being on block entry the coolant always enters the block at the same temperature but with extra cooling at the radiators the thermostat has closed more to slow the flow rate to match the additional cooling available.

Because the coolant travels slower through the engine the exit temperature is higher – unless the two radiators are not coping in which case the thermostat is fully open and then the extra radiator will provide additional cooling.

In any closed loop system that has had time to settle down to a constant temperature you cannot take more heat out at one end of the system without putting more heat in at the other and if that “more heat in” starts at the block entry thermostat then it can only absorb more heat on the way through the engine and exit it hotter at the top (where the cylinders and heads) are and it is beneficial to keep them cooler to make sure the higher temperature does not make the oil in the cylinders go too thin – so it keeps a good oil film thickness keeping the piston as far apart from the cylinders as the bore clearance allows.

But - Running a LTT at the engine entry will always lower running temperatures (unless 2 rads are not coping).

To make sure the 3rd radiator only comes into operation when it is needed - you need a thermostat at entry to the third radiator (so it only opens when the other two radiators cannot cope i.e. set at a slightly higher temperature than the main thermostat). A manual third radiator blind could achieve the same objective.

We went even further on our 996 3.4 race engine when we removed the entry thermostat and replaced it with an exit thermostat housing that had two entry points (one for each cylinder block) and two thermostats in “Y” formation feeding into one outlet pipe – as this automatically balanced the exit temperature for each cylinder bank. We then fitted another thermostat to the third radiator inlet set at a higher temperature so it was only open when needed.

We also always alter the feeder slots to the block to increase the block flow compared to head flow (you only measure the combined temperature of the two).

It is not rocket science although I appreciate that it does seem odd that an additional radiator can make the internals of the engine run too hot – but it can AND DOES.

We made (and stock) a thermostat housing and thermostats which owners can fit with their own pipework between the radiator and the third radiator for earlier 996 models, but struggled to make a commercially viable one for the 987/997 due to its complex plastic short hose between the radiators and the radial locator and clip design.

The 996 one could be used if the owner is prepared to cut the pipe connecting the two radiators and fit two bends and additional pipework to fit the 996 earlier housing in a loop (that might also require repositioning a horn).
We were well on with designing a 3rd radiator thermostat housing when Covid 19 hit us and we had to make a number of changes to our machinery layout (because we had 3 people in close proximity in the same congested machine shop – which was then illegal).
After creating two machine shops and investing in more machinery (to try and maintain efficiency) we then had delivery and staffing problems (with self - isolations) and of course had to prioritise production to customer engine rebuilds ( to try and maintain profitability) and shelve small batch production of thermostat housings in tiny quantities (that will never be a volume earner nor probably profitable lol – things like that we make for racing also are into small a quantity to make profit and are regarded more as a service to our racing customers).

Meanwhile the performance of our capacity increased models also increased our workload making it difficult to turn off production machinery to stop and make a few thermostat housings.
Then the improved economy of those bigger engines resulted in us turning our attention to the development of our Eco-Power engines (an idea that will cut down green house gas emissions) which we regarded as more important and responsible than turning attention to 3rd radiator thermostat housings.

However - we did end up with a solution to produce them – but it will take some tooling and time to produce a batch which right now is too far away to confirm availability.
So I apologise if people were waiting for that product and we failed to produce it sooner. I hope we will get round to it sometime soon.

Baz