Thermostats. Yeah, I know. But why?

[jbcollier]

Crossflows in Lotus Seven S3s did not have thermostats fitted.  Supposedly if you fit a thermostat they will overheat.  I'm a retired, 35 year mechanic and I like thermostats.  They provide quicker warm up and better heat dispersion while doing so. Can anyone confirm this is actually true?  If so, any idea why?

[jbcollier]

[SENC]

Precrossflows did not have then either, at least not those with swirlpots.  That didn't stop people from  adding them, and mine had one between swirlpot and head when I got it.  I could never get that connection not to leak (gaskets, wellseal, hylomar) with the thermostat and, seeing no particular value to having it in my warmer climate I ditched it.  I'm careful to let it warm before pushing it much, but it doesn't take long.

[jbcollier]

The S3 has a swirl pot hooked to a standard thermostat cover by a hose.  Sealing should be fine.

 

Did yours run hot at all with the stat in place?

[IamScotticus]

I suppose opening the heater circuit or unplugging the intake manifold coolant circuit and pluming into the water pump could help alleviate hot spots while warming up.

 

I suppose one would need to know what the running trmp is without  a thermostat. 

Edited October 20, 2022 by IamScotticus

[SENC]

Not really, but I didn't put a ton of miles on it before removing the thermostat.  That said, as I was attempting to solve the leak with the t-stat I remember thinking that the thermostat was a restriction in the system even when open (due to the cooling system design) and wondering if that restriction would impact how the swirlpot and system overall functioned.  Mine comes up to temp reasonably quickly and then holds steady in the desired range very well - I only turn on the (small original electric) fan when I'm sitting still in traffic on a hot day.  In that latter case I do need to blip the throttle periodically to get it above 2k to make sure the pump is pushing water, but I haven't had the temp fluctuations I sometimes experience in my Elan.  It doesn't take much forward movement and natural airflow for temps to return right where I want to see them.  Unlike the need to worry more about temp differentials due to our aluminum twink heads in the Elan and Europa, I think the precrossflow and crossflow heads heat up more evenly with the block so suspect there is a little less to worry about - but I'd defer to your better background and experience whether I'm just being hopeful.

[jbcollier]

I never discount actual, practical, experience.  Thanks, I appreciate your comments.

[IamScotticus]

Might want a restrictor in place of a thermostat to slow down the water flow.  

[MV8]

Their recommendation to leave it out and why makes sense but we have better options these days.

The faster a given amount of coolant circulates, the more efficient the cooling system will be (i.e, more radiator passes in a period of time for the same heat load).

If the coolant flow is too rapid for the passages, destructive cavitation can essentially erode the hard parts like the pump body, timing cover, impeller, etc but the stat is not the most restrictive element so I don't see the harm in that except for potentially slower warm up with a substantial cooling system. Ambient would also be a factor.

I am not suggesting that anybody do this, but the greatest restrictions are the holes in the head gasket. In general, these are typically enlarged for endurance racing with pump and passage improvements for higher flow. A tiny change makes a big difference.

There should be a bypass passage other than the heater circuit that can be completely closed. This can be a drilled stat, sometimes a small hose near the pump, or an internal passage or poor fitting stat. The bypass prevents hot spots during warm up. The amount bypassed is directly proportional to the level of numpty behind the wheel. Old cars with carbs have a built in safety of not running so well cold. Modern cars don't stop a person from abusing the cold equipment. All bypasses bypass the radiator and provide no cooling.

I generally restrict the bypass, run a 180 stat, ensure the rad has baffles, and fit the biggest electric fan with the most amps I can find with a toggle and a thermo switch relay control.   

[IamScotticus]

Just for the sake of discussion,  isn't it a fact that coolant flowing slower with higher pressure will absorb more heat than a faster coolant flow?

[MV8]

There are a number of factors that make up what is a net improvement in heat transfer.

The amount of heat transferred is directly proportional to the mass flow, temp delta across the exchanger (coolant versus air), and the time frame across it.

If the pump pressure is increased (with design improvement or pulley ratio changes) or the restrictions reduced (staying out of cavitation), the mass flow will increase and the pressure will increase at the exchanger inlet.

The cycle time of a drop of coolant across the exchanger is reduced but there are more passes for a given period.

The temps across the entire system will be more uniform as well.

[SENC]

Agree with MV8, slower flow allows more heat transfer per unit of liquid... but increased flow means more of those units are passing through, so total heat transfer (generally speaking) increases with flow.  A limitation to this, as MV8 mentioned, is when flow reaches a point it causes cavitation.  There was a great discussion on this on lotuselan.net a few years back that I refer to periodically, I'll find and link.

 

In the particular case of my precrossflow 7, as it was apparently designed for use w/o a tstat I adopted the opinion that operating without would not result in cavitation.  So far, lack of overheating whether at long upper-rpm speeds or in low-rpm traffic suggest it is ok.

 

https://lotuselan.net/forums/viewtopic.php?t=29674&f=39&start=0#p197209

 

and an on-point excerpt from that thread, if you don't wish to read the entire thing...

As they say in the mythbusters - "warning a little science needed here"

Heat transfer rate = K x A x LMTD

K a nominal constant but determined by a number of geometric and surface condition factors including "Reynolds number" which is a measure of turbulence in the flowing fluid
A the heat transfer area
LMTD is the log mean temperature differential

You are correct AHM in that there is no specific mass flow term but there is "Reynolds number" this typically increases with increasing fluid velocity increasing the K factor. Thus increasing mass flow through a heat exchanger typically increases heat transfer coefficient but the rate of change is nomrally small . The big effect as you observe is the increase in what is technically known a LMTD = "log mean temperature differential" as increasing flow means the flowing fluid heats up less as it pass through the engine and thus picks up more heat from the engine as there is more temperature differential driving the heat flow. It picks up less heat per unit of fluid flow but more fluid flow so overall more heat transfer from the engine, I think it is this complexity that people somethimes get mixed up with.

In an engine you have two heat exchangers in series - the first is the engine sending heat to the coolant and the second is the radiator sending the coolant heat to the air. This makes the overall situation a little more complex but in general a faster coolant circulation will tranfer more heat --- which is why the thermostat opens to control the engine temperature.

Removing the thermostat could cause a problem with the water pump cavitating and flow dropping off rather than increasing but i dont think this is a problem with the twin cam. Removing the thermostat reduces the pressure differential through the heater core so it will reduce the heater effectiveness. It also means a longer warm up time which is not good for engine wear. it may also mean the engine runs so cold that blowby contaminants stay in the oil and dont boil off leading to oil contamination and engine wear due to this also.
cheers
Rohan

Edited October 21, 2022 by SENC

[jbcollier]

For a detailed, practical analysis, I found this one to be most interesting:

 

 https://www.teae.org/cooling-the-sunbeam-tiger/

Edited October 21, 2022 by jbcollier

[SENC]

Good read JB, thanks for sharing.  Practical application and testing like this is always valuable to me in understanding how things work.  The primary focus on airflow as a priority definitely rings true to me.

 

The water pump and flow testing was also of interest, though I would have liked to see them measure flow in some way as they measured air flow.  In Elan twinks, it is well known that the water pump simply doesn't push enough water at idle for good circulation (the same is surely true for our early Sevens) - that it really needs to be over 1500 rpm for sufficient flow to cool.  Many have solved this issue with electric water pump conversions, some with smaller pulleys to lower needed rpms.  I haven't changed either, instead relying on throttle blips if stuck at idle for any period.

[MV8]

1 hour ago, jbcollier said:

For a detailed, practical analysis, I found this one to be most interesting:

 

 https://www.teae.org/cooling-the-sunbeam-tiger/

It looks like TEAE went to a commendable amount of effort to perform a wide range of tests in a short period of time, but a chart with no supporting data is easy to debate, leaving more questions than answers but FWIW, I agree with most of it. I'm no stranger to trying to cool an SBF in a confined space.

Police cars typically have mods that include restriction washers in the bypass, contrary to the tiger findings for the same family of ford engine, as does oem documentation (gm) on modification (of gm engines) for endurance racing.

[ashyers]

One item that I don't believe mentioned here is the local pressure in the cylinder head pre-Tstat / Restrictor. Often the local pressure in the head will exceed the cap pressure. This helps prevent localized boiling around the exhaust valves which can cause all kinds of fun issues.