MG-Cars.info
MG MGF Technical HGF Soln... Image Needed BBS discussion at MG-Cars.info
MG-Cars.info
Welcome to our Site for MG, Triumph and Austin-Healey Car Information.
Parts
MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.
|
|
MG MGF Technical - HGF Soln... Image Needed
| Hi, I've finally got around to typing up my analysis / root cause for the HGF. Im creating a document detailing the cause of the failure and the soln. I need a picture of the coolant system to carry on... Can someone who has the RAVE disc handy, send me a grab of the image that shows the coolant flow directions in the system. My copy is at work. If there are other handy looking images of the system It will make writing my report easier. (I had a £10k datalogger hooked up to the car pre & post fix and have evidence that strongly supports the theory and solution.) Dave |
| David Monks |
| http://www.mgf.ultimatemg.com/group2/coolant/index.htm If of any help. |
| Mark.L |
| Right. Thank for the link - the picture is perfect. I have written the report now. I will proof read it tomorrow and post a link to it accordingly. Dave |
| David Monks |
| Now this might be an interesting read. |
| Yigit |
| Can I reproduce it for the website Dave? :o) |
| Rob Bell |
| I will post the document location this evening - just finishing the proof reading now. Its a large document, 20sides, and will eventually make a short 1side abstract to send to the MG mags. Rob, reproduce as desired. I have used some info's from your site anyway (along with others) so you are credited at the end. Dave |
| David Monks |
| David, does this work include proper solutions to the HGF issue? |
| Neil |
| Thanks David - I look forward to reading it :o) |
| Rob Bell |
| >>David, does this work include proper solutions to the HGF issue?<< It doesn't include anything else! Dave |
| David Monks |
| I'm all excited, do we do away with the crazy jumble of hoses around the engine? and let gravity assist rather than work against us in the rad? Or is it as easy as to simply induction weld the head on? Maybe a Papal blessing? Come on, post it, we'll do the proof reading:0) G |
| g hampson |
| I have now completed the document. Similar solutions have been used before, but I believe a scientific comparison of pre and post modification has not been conducted before. Comments and questions are welcome - I may have forgotten to include something! http://www.mgfmavhh.ukf.net/ Dave |
| David Monks |
| Well researched! Very interesting temperature readings. What you are demonstrating has been said many times before, but this puts a bit more science to the problem. It may be that this is a better "fix", and thermally it should be. It may also mask any mechanical intolerance in assembly, ie liner heights, head shuffle, oil temperature etc, but this level of block/head temperature control is sufficient not to call any discrepancies of production fits into a significant role. Maybe? Worth trying. Is the TF PRT a suitable unit to fit at the front? It would certainly make a more "standard" fitment, and the longer length of under floor pipework would act to improve bypass flow, whilst acting as a damper on sudden temperature fluctuations whilst warming up. Got to think about this now, as I was about to fit other variation soon to my car! |
| Jerry Herbert |
| DAMN I havin the car serviced next month remote thermostat will be fitted... well done sir. Alex.. |
| alex |
| To be honest, Im not sure what to think about the PRT... it seems like an unneccessary complication to me. The 'traditional' way to do it seems the most sensible to me. If someone has a PRT fitted, and an unnatural desire to be a mad scientist, I'd be willing to instrument the car up and run a comparison to my fix. Dave |
| David Monks |
| I think if you fitted a TF PRT at the front, but run backwards to normal flow, then the interconnecting bypass would be progressively shut off - an advantage? - giving full flow through the radiator at max water temperature. Something to think about tonight!! |
| Jerry Herbert |
| David thats some great work you've done and the results look good. A couple of small points though IIRC from the KingK article the thermostat arrangement on the K is not unusual for this construction of engine as I think Renault used the same for their alloy block wet liner engines, also "The bulk of the cold water, i.e that in the radiator is at ambient air temperature. The radiator is not heated by conduction or radiation as it would be in a front engine vehicle", I can't really see how conduction will play any significant roll in heating the radiator and less still radiation as the engine is at maybe 100C so will be insignificant. Even with the exhaust nearby and cherry red the heat radiated would be have little effect given the cooling airflow. |
| David Billington |
| Okay, this point can be debated. When we're only talking about 100oC operating temperatures, the radiator being 10oC warmer is a significant difference - surely the radiator in a front engined car is hotter more quickly due to the close proximity to the engine - conduction via the stationary coolant in the pipes? You can see the 12oC temp drop along the underbody pipes in my data with moving coolant - so the effect must be there more significantly for stationary coolant. I take the point about other engines possible using this same setup - i cant be certain because I know very little about non-rover & non-bmw engines. However, the thermal data cannot be argued with - the thermal shock effect is removed by moving the thermostat to the 'traditional' side. Dave |
| David Monks |
| Interestingly you mentioned Renault engines using a similar system... What Car?: "Renault Mégane Saloon (96-03) RELIABILITY. Holds together reasonably. Some have had head gasket failure...." Honest Jon: "1.6 8v prone to head gasket failure." "Low-mileage cars had head gaskets replaced FOC." Maybe a reflection of the 'copy cat' nature of the motor industry? |
| David Monks |
| David (Monks) so people don't start thinking i'm writing to myself. I agree that the difference between say 3C and 13C may be significant with respect to 100C but am not sure that the conduction through stationary coolant will be significant in the time frame of the engine heating. What may be significant is the cooling you noted along the MGF pipes as the thermostat starts to open and provide some hot coolant into the radiator circuit, from your data it would seem the coolant returns probably cooler than for FWD so exacerbating problems in the standard thermostat system. I agree though with your conclusions that the lack of proper blending of cold coolant with hot will lead to thermal cycling. I wonder if the external type thermostat often used on FIAT/Lancia twin cams might be of use. This was plumbed to circulate water about the head until warmed where it would then mix the flow with coolant through the radiator, this would transition from no rad flow to full rad flow as the coolant temp increased. Re the Renault a quick look at a R25 manual showed thermostat regulation on the input to the block for the inline 4 but apparently on the outlet for the V6, both wetliner engines and predating the K by some time. |
| David Billington |
| i've only had a breif look at your graphs, but isn't the effect shown just due to higher water flow in the bypass causing the diffrence in coolant in/out to reduce and nothing to do with thermostat position. the only reason for the thermostat to be moved being that there isn't room for a high volume bypass in the traditional position |
| Will Munns |
| Well done ;) Great works and approved theories. Carl will like it ... scroll down to the bottom. http://www.mgfcar.de/thermostat/Water_System_Mods.htm I've put a link there to your site. |
| Dieter |
| Will, the thermostat being at the front is CRITICAL. The volume of cold coolant in the system MUST be much less than the volume of hot coolant. - otherwise the blending effect will not work. Just adding a high flow bypass tube does not reduce the volume of cold coolant. Dave |
| Dave Monks |
| given the same coolant speed and the same heat to dissipate, the temp gradient across the head will be the same regardless of where the thermostat is. Incressing the bypass size incresses the speed of the coolant at all times, therefore reducing the max gradient but reduces the max cooling capacity of the engine. At the point in the system of the thermostat the temp is clamped, if you clamp the inlet temp then the outlet temp will vary, clamp the outlet temp and the inlet temp will vary. if the bypass volume is low then mixing will use little radiator coolant and the gradient will be higher, volume high the gradient lower If the outside temp is low then a traditional thermostat will cause the volume of coolant to be low = high gradient If the outside temp is high then a high volume bypass will cause the thermostat to be unable to mix in enough cold water to cool the engine. If a high volume bypass is used then the bypass must be progressivly shut off at higher bypass temps to ensure that the coolant speed and thus gradient remains the same. |
| Will Munns |
| Following this with interest, as I'm about to do something myself on an 02 TF. I'm not an expert, but-- I'm not sure I follow the last comment! If the remote thermostat is at the front, with a large by-pass, the volume of water warming up will be larger. Warm up time might be longer, but more controlled because of the bigger mass, and the oil will have more of a chance to warm up, matching the water. Yes, it does seem strange having all that water running up and down the car as a by-pass, but that might be all that is required for short term low speed driving. More energetic use will bleed in cooler water with a long mixing passage. Using a TF PRT in reverse flow to normal will seal off the by-pass when at max temperature? Or am I missing something? Probably! |
| Jerry Herbert |
| Excellant and to my mind, pretty obvious, not the method you've come up with, but the fact that more constant temps = less stresses = less movements = more stable sealing - irrespective of the actual mechanisms facilitating this. I have a couple of questions which I am unsure of in your paper. (once you put pictures there i loose the ability to read) Is the original 'stat removed entirely? I know this is the standard set up but since I first saw the layout of the cooling system the following has always puzzled me - The top hose is the cold water out of the rad - why? cold water drops, does this arrangement not mean you'd have a large dead area in the rad where the pumped water flow and thermal flows are working against each other? is there any benifit in switching the two? maybe you could get your toys out again and see? Thanks Gareth |
| g hampson |
| The temperature gradient across the head is not really the issue - it is the CHANGE in the temperature gradient. The cold water being allowed into the head causes a sudden drop on the inlet side - ie the gradient suddenly increases. This can only happen if there is a lot of cold water in the system waiting to get past the thermostat. With the thermostat at the front, the cold coolant 'reservior' is a lot smaller - there is also a greater opportunity to blend with the hot water from the bypass, ie the cold water does not get to the head. This setup regulates the temperature of the engine, the original setup regulates the temperature of the radiator. A higher flow bypass on the original setup would help - but the effect can not be as significant - there is only 5cm of 'blending' space left - the cold water still hits the head. Refer to the last graph 'head temp differential comparison' - this is the crux of the matter. The dip in the redline at 550-800sec is the cold water surging into the head - of course the gradient eventually settles out, but the damage is done - and like all fatigue stresses, the effect is cumulative. The blue line, ie the new setup, completely erradicates the sudden change in differential - ie the stress is eliminated. A spin off is the differential across the head is, on average, much lower - so again, the torsion forces across the head is also eliminated. Answer: Yes the original thermostat is removed, infact, it is replaced with a 'blank' which is made by cutting the centre out of a scrap thermostat. This is necessary to maintain the water tight seal in the 'disused' thermostat housing. Dave |
| Dave Monks |
| Top hose / bottom hose issue: To be honest, im not 100% certain here - if your talking about a stationary tank of water then cold sinks to the bottom... But the radiator is a flowing set of horizontal veins, so not a tank, and not stationary. It would be interesting to measure the difference in swapping the hoses top-bottom... but a lot of work for not a lot of gain, maybe?! Anyone know any better? Dave |
| Dave Monks |
| Warm up time issue... I too thought that the warm up time might be extended a bit... but its not significant - look at the graphs in the 'pre-post modification comparison' - both setups reach max temp at about the same time. Using the PRT to seal of the bypass introduces a new set of variables, I can't imagine it would do much harm, but it seems like an unnecessary complication and expense - why bother... it doesnt need a PRT. This setup can be built for under £20! A PRT alone would cost more than that! btw your comment about the 'long mixing passage' generated by using the thermostat up-front is spot on - this is exactly what I was trying to achieve. Another nice feature, is you can get to the thermostat VERY easily - unlike the original one... I was asked if I 'self harmed' the following day at work, I still have the scars from getting that bloody thermostat out!! Dave |
| Dave Monks |
| FWIW: This is a cross-flow radiator, not the "old fashioned" vertical tubed sort. The headers are vertical, I guess the hot inlet is at the bottom to stop it staying at the top and going straight through, reducing effectiveness. Yes, the front position is easy to get, saves disturbing the crowded engine compartment. The PRT will shut off the bypass only when full temp is reached, ensuring all the hot coolant enters the rad when needed most. I don't know the cost of a TF PRT, but it would eliminate by-pass bleed when hot, and mimic the original PRT function, but located in a better position!! I think? |
| Jerry Herbert |
| The radiator pipework puzzles me as well?. Because of the front radiator, rear engine set up of the F/TF you loose some of the Gravity effect you get in the normal front engine/front radiator set up of most cars, ie top hose to top radiator connection. However once the hot coolant enters the bottom of the radiator (as it is in the standard set up), it will rise quickly to the top of the rad, thus leaving the bottom of the opposite entry side cooler, therefore not using the full surface of the rad, which a normal hot in at the top, cold out of the bottom set up would do, ie the cascade effect- hot water falling across the full length of the rad as it cools!!. Doubless there is a good reason why Rover did it this way......... just cant think of one! :o/ |
| Mark.L |
| thermocycling wont happen in an k whichever way up the radiator is, that is because the thermostat is on the inlet, so when the engine switches off the water moves very slowly, so the very cold water from the rad will cool the thermostat and shut the rad off immediatly, put the thermostat on the outlet and the very hot water from the engine keeps the thermostat open _because_ it is flowing slowly. Given this there is little point rover trying to make the system self syphon, so ease of parts supply becomes the only factor. |
| Will Munns |
| BTW Rover sell a thermostatless thermostat for use with the PRT thermostat |
| Will Munns |
| The differential peaks on your 'after' graph look about the same hight as those on your 'before' graph, the only diffrence I see is that the differential over the block as a whole is less and this could be easily explained as a double size bypass (you now have one in the engine and one at the front). BTW, I think this is exactly what this board is for - even if I don't agree with your conclusions you are bringing things to the wider world, with real data, so please don't simply dismiss me as a refusenik, there is a problem and there are various 'solutions' from a number of people. What would be really great would be to do the same thing on a pre and post PRT instalation - but I guess you are like me and would rather take bits from a scrappie and make your own than put money on a till and take someone elses design! |
| Will Munns |
| Will, The differential peaks are about the same when the engine is "at temp" But with the mod the large initial cycle differential (which I guess is occuring as the 'stat intially opens and then quickly closes due to the volume of cold water in the system) is removed? Do you not agree it is most likely this initial "shift" is the major cause of "shuffle". Also the reduced overall running temp would limit, to some degree at least, overall movement.? |
| g hampson |
| I think that the design of the engine should be put in the secton on causes of HGF. The very narrow lands of material for the gasket to contact must be a factor in reliability. I was amazed to see how little distance there is between the water passages and the cylinders, and indeed between the cylinders themselves. A very interesting write up David, thankyou. Bruce |
| Bruce Caldwell |
| Exactly! The initial differential jump is huge - between 550-800secs if i remember correctly, this is totally eliminated on the 'after' graph. The cycling beyond this is only small in both cases - but the initial thump is repeated each and everytime you start your car - the cumulative effect is a fatigue failure. If you only want to avoid 'bypass bleed' - why don't you just use a normal 'valve' type thermostat (eg Rover 620TI) - what does the PRT do that is so favourable? To be honest, it dont think it matters much, the point is the timing of the opening and closing of the thermostat - not the method by which it is done. >>dismiss me as a refusenik<< of course not - this is why I posted this here! I want to avoid HGF as much as everyone else - I have posted here for constructive critisism. Dave |
| Dave Monks |
| Any idea what an ideal bypass size would be? Too small and there will be no 'blending' when the thermostat opens, too large and the radiator will not have enough of the flow (think ahead to those barmy summer afternoons, hood down, country lane...). It may not be very critical, but would it be possible to shut off the bypass at a higher temperature when radiator cooling is more important then blending. Maybe someone should invent a more gradual thermostat that slowly allows more water to flow as the engine warms up so all the water has had time to warm up before its fully open. John. |
| John B |
| Right! now let's get serious! Enough of the verbage! Turn this into a good value kit for us all!! |
| Neil |
| I think the point of the 'verbage' is to accurately assess the problem and consider all variables before proceeding. |
| mike roberts |
| >Exactly! The initial differential jump is huge - between >550-800secs if i remember correctly, this is totally >eliminated on the 'after' graph. is this not a function of a worn out thermostat? the thermostat is the item in the system that is not reacting fast or smothly enough, so it tends to stick and then overshoot, and then stick and then overshoot, once it has moved in and out a bit it is freeer and the rest is just hysterisis in the system (which you still have). I would expect every thermostat to do this dependant on age an manufacturing quality rather than position in the system. I think the 'volume of cold water in the system' is a red herring, but there could be some mileage in the long pipework back to the engine heating up and therfore damping the osscalations, but I would expect this to show up on the second graph as reduced deltas late in the trip, which are not there. >>dismiss me as a refusenik<< of course not Good, because I am here in order to talk this thru and come up with the best solution for my own car, and I may have faulty thinking too! Did you arrive at the conclusion before starting the tests, or were you looking for a specific pattern? |
| Will Munns |
| Well, I've just bought a PRT this morning: part no PEM101020. I've convinced myself that the position of the PRT near the radiator will give more thermal stability to the cooling water. Looking at it, I'm going to run the PRT backwards to normal flow, that way the by-pass will be blocked off at max thermostat opening, replicating what it does in its normal place. Been looking at the temperature of the various pipes by hand(not very scientific), but I'm sure the smoothing of the thermal hysteresis will tend to mask any production tolerance discepancies, which might lead to problems. In any case, I think there is nothing to lose by trying this modification (except the cost of the PRT!) |
| Jerry Herbert |
| Dave, Excellent work! I assume your F has an early coolant system. It would be interesting to see the readings from a MGR PRT post 2002 model to compare the two. One assumes that MGR engineers must have done a similar exercise and the PRT was their solution to problem(?) Have you any plans in this direction? Regards Graham |
| GrahamH |
| >Well, I've just bought a PRT this morning: part no PEM101020. where from Jerry? |
| Neil |
| David Good to see an attempt being made to get to the bottom of the issues in a rational manner. The questions that occurs to me are: 1) Is there any proof from elsewhere that an operating range of 10C degrees is sufficient to ultimately cause movement of the head/gasket? Really to complete the data you need to look at another engine configuration know to be satisfactory, eg a front engined MG (ST?). At present we imply the problem has its root in the low volume of coolant in the engine compared to the rest of the circuit, but there is only circumstantial evidence for this, because the Freelander is also weak. 2) Are there any other factors that might have a greater effect than temperature? I am still intrigued by a statement made by Roger Parker some years ago when he used a strobe light to see how much the engine block distorted when running. This seems to me to be far more likely to cause failure - but again I have no proof - and it would be much more difficult to instrument. Chris |
| Chris |
| The thermostat was not old or worn out. It responds perfectly to hot / cold coolant. In the original setup there is enough cold coolant available to close the thermostat immediately after it opens. The volume of cold water is NOT a red herring - it is at the very heart of the issue. Temperature changes is what we are trying to avoid - this can only be achieved if 'on average' the coolant temperature is regulated - if there is a large amount of cold coolant not involved until the thermostat opens, then there is the capacity for large temperature change. As for the position in the system - this is also of utmost importance. It is in the wrong place, the logic behind fitting it there is flawed at least - the results of this mistake are plain for everyone to see. Other engines with this setup also suffer from premature HGF. It should be on the other side of the engine, and in a rear engined car, it should be further away from the engine. >>MGR engineers must have done a similar exercise<< I wouldn't assume that - they had very little money to play with, and secondly, a re-engineering of the engine would be far to expensive - whatever solution they went with would almost certainly have to be a compromise. I am a scientist and an engineer - and regardless of what my 'gut feeling' was on the issue the conclusion of the testing has been backed up by rational explanation. ie I did not set out to proove my gut feelings correct - that wouldn't help my cause much. Dave |
| David Monks |
| >>because the Freelander is also weak.<< The freelander is subject to another issue - it is underpowered... ie the engine is always working hard, so it runs hotter etc etc |
| David Monks |
| >>>MGR engineers must have done a similar exercise<< I >wouldn't assume that - they had very little money to >play with, and secondly, a re-engineering of the engine >would be far to expensive - whatever solution they went >with would almost certainly have to be a compromise. perhaps, but what you have done is not novel (the elise boys have been talking of it now for at least 4 years) and it doesn't look expensive, at least not in the scheme of a company doing the research, and it would be no more expensive to do than fitting a PRT, with a very marginal cost of some extra pipework. They clearly _did_ do some research to come up with the PRT, a compromize it may be, but your testing was donw with a much smaller budget and only one car - so I wouldn't dismiss them quite so quickly. >The volume of cold water is NOT a red herring - it is at >the very heart of the issue. Temperature changes is what >we are trying to avoid - this can only be achieved if >'on average' the coolant temperature is regulated. On a cold winters day on the motorway the coolant comming back from the radiator is just as cold, the issue is how it is blended at the thermostat, a jerky thermostat will give you high deltas, jerky can be the cause of a number of things, one of them could be a large opening gap at small temp changes, this would lead to a high hysterisis due to overshoots, but you have to ask yourself why you see this at the beginning but not during normal running. |
| Will Munns |
| Got my PRT from a local ex MG-R dealer. £24.29 + VAT. The rest of the bits I have anyway. Got to agree with David M so far, logic, sense of "rightness", etc tells me the cooling system is going to be more free from sudden temp changes. If this only hides poor liner heights etc etc, then it is worth a try. Even if it prolongs the inevitable, I'll stick with it as it feels more right than the non PRT set up I've got now (02 TF) |
| Jerry Herbert |
| >>perhaps, but what you have done is not novel<< I know, I stated that at the start, but this is the first time the car has been 'proffesionally' instrumented and root cause analysis conducted. It is also the first time the 'solution' (however 'novel' it is) has been checked and proved with appropriate equipment - not just guess work? or am I wrong? >>but you have to ask yourself why you see this at the beginning but not during normal running.<< Thats the point! The initial 'bump' is caused by the large volume of cold coolant in the system - after that it has been through the heating cycle once or twice and is now much warmer. ie the differential change is much smaller. Secondly, the liners all start off at the right height, the pump starts off working correctly, the bolts are tight and everything else is fine... it is the repeating cumulative effect of the thermo-mechanical shock that causes these parts to move/twist/stretch etc etc. These are the cause of the HGF... but not the absolute root of the cause. The absolute root cause is the cumulative effect of thermo-mechanical shock / fatigue on these parts. Dave |
| David Monks |
| HELP!!! What is difference if any between PRT and remote thermostat?? I just want to get the right item for my car which lives in London mostly in traffic. Also before they sell out after the discussions going on about them!!! |
| kate |
| >after that it has been through the heating cycle once or twice and is now much warmer i'm not sure this is correct, on a frosty morning cold motorway the air is whipping thru there at 70 miles per hour, I would be very surprised if the coolant was any warmer on a typical day |
| Will Munns |
| Having being flamed a while back for suggesting the design of the engine to be flawed on the basis that the material of the bolts expands at a differing rate to that of the block and head. I have therefore watched this thread with interest and it was your last comment David that caused me to post. Rightly or wrongly it was suggested that the bolts are STRETCH bolts and as such can keep the same amount of effort holding the engine together whether hot or cold. You appear to be doubting this theory by suggesting this constant stretching will inevitably lead to fatigue and failure. I hope you are correct in your assumptions and can find a solution, it will be very interesting because there is no doubt in my mind, that along with high pension payments warrently claims must have had a serious affect on the profitability of MGR. Why their engineers and budget failed and yet you believe you can solve the problem will be interesting to follow. |
| Bob (robert) I am coming out!! yes I once owned an MGB!! |
| >>Any idea what an ideal bypass size would be? Too small and there will be no 'blending' when the thermostat opens, too large and the radiator will not have enough of the flow<< Sorry, I missed that up there... The bypass is 'throttled' by a 4mm hole in the sleeve to the bypass tube. Bypass flow doesn't need to be too large for reasons you state above. The pipe itself is about 12-15mm id. The flow should be split in the ratio's of the area's of the pipes... Without wanting to revert to Reynolds / Navier-Stokes equations, I tried the following: 1. Cold engine running - rad pipe soft, bypass section harder... warms up, until the stat starts to open... 2. Once the stat is fully open and the engine is at full operating temperature, the bypass tube is no longer hard - and the rad hose now is. Seems to me that the bypass is 'bypassing', 'blending' and 'redundant' in the correct circumstances. The radiator is obviously working correctly: since the logger has been removed I have done several 300miles 'round trips' without any cooling issues at all. >> ...bolts are STRETCH bolts and as such... << I don't think the bolts stretch. The bolts are known to loosen, so either, the bolts are stretching (which they are, but they recover), OR, the head becomes compressed, the gasket becomes compressed, the deformations push the liners down... or of course a combination of all of the above. "aluminium has a tendency to "creep" under steady sustained pressure (to a greater degree as the temperature rises), again producing a degree of looseness in an initially tight connection." -Wikipedia. Dave |
| Dave |
| "Stretch" actually not my words Dave but the words of some members of these boards in defence of the design of the engine. I have to admit at not accepting this concept as being any good although I would need to bow to the designers better knowledge of such things |
| Bob (robert) I am coming out!! yes I once owned an MGB!! |
| "I don't think the bolts stretch. " that was supposed to read like "they dont PERMANENTLY stretch" Dave |
| Dave |
| the k manual calls them stretch bolts, the idea is that they are stretched during installation and they _will_ have different loads dependant of temp because the alloy expands at a different speed to the steel, but the same is true of the liners and on all alloy engines- not just the k |
| Will Munns |
| Good work and most interesting indeed. Now, is the PRT a definitive solution to the HGF? As I am obsessed with this potential trouble -my F is 33K on the clock- I recently adressed the MGF Centre asking for advise about the HGF subject. They suggested that even with PRT, K engines often fail, and suggested an oil cooler as an eventual prevention. Mike |
| JM Vega-P. |
| an oil cooler?? sounds like they have a job lot in! nope the PRT does not -cure- the problem, just makes it less lightly |
| Will Munns |
| Hey guys, I agree with you all!! I think this is a good shot in minimising the risk of aquiring HGF. Worth a try in any case, as I can see the logic is sound. But engine developement is not going to happen with us all messing around - money needs to be spent. It may be academic for me now, as my B&G warning sounded off a few days ago, and I seem to be slowly losing coolant with no obvious leaks. Just as I had rushed out and bought the PRT today!!! |
| Jerry Herbert |
| Jerry, why not mount your oil/water cooler up front - it has a thermostat built in as I recall? The only disadvantage is the long oil-lines - but that's a low price to pay for the packaging ease. David - thanks very much for posting your data - sorry I am coming to this conversation so late! The fundamental flaw in the original cooling system was the flow rate through the bypass pipe - and this relates to your thinking surrounding the remote location of your thermostat solution. In essence, the problem is that the flow through the bypass pipe is inadequate. If you could up the flow rate through this pipe, you wouldn't need to relocate the thermostat at all - notice how the bypass outflow mixes with the radiator return flow. If the bypass flow rate were high enough, there would be adequate blending of radiator return and bypass to prevent the oscillation in thermostat function. The practical problem is to how to achieve this high flow with the thermostat in the standard position, and then how to avoid the pitfalls that Will points out - namely inadequate cooling under extreme engine loads in high ambient temperatures. One thing that does strike me, looking at the temperature graphs - and that is the temperature overshoots on initial warm up. Really gunning the engine from cold is really going to exacerbate that particular problem - and just goes to show what we've all been saying all these years - be sympathetic on the engine until it has had a chance to warm through! Regarding the PRT - I'd love to see how that works too! There are days when I think I understand what it is doing, and days when I don't... Would like to be able to see the raw data! David - I've dropped you an email about another cooling system project... ;o) |
| Rob Bell |
| >> that is the temperature overshoots on initial warm up.<< Rob, which graph are you looking at?... I'll comment when im sure what your refering too. Re bypass flow, im not sure JUST making the bypass flow greater would help much - you are still left with a lot of cold coolant in the system that can 'dilute' the warm bypass flow when the thermostat opens. >>inadequate cooling under extreme engine loads in high ambient temperatures.<< I don't believe the new cooling system will underperform - infact, I think it will be more effective than standard. Dave |
| Dave |
| just been offered a PRT and all the gubbins from MGF centre for £250+VAT Ouch I say! What if I just drill some holes in the existing thermostat, would this not reduce the thermal shock issue? |
| Neil |
| Dave, http://www.mgfmavhh.ukf.net/index_files/image041.gif shows the difference between inlet and outlet temperatures, and shows what happens in greatest relief (although the same overshoot is evident on the thermostat housing traces). It is interesting to note that with your modified remote thermostat solution this is significantly attenuated... I suspect that you could have made that standard graph look considerably worse than it is by burrying the throttle from cold if you wanted to - which would have been interesting to look at, just to see how bad you can make it! >> Re bypass flow, im not sure JUST making the bypass flow greater would help much - you are still left with a lot of cold coolant in the system that can 'dilute' the warm bypass flow when the thermostat opens. << Ah, but it will - the unit volume of bypass coolant to unit volume radiator return coolant per unit time will be higher, meaning less dilution. Extrapolate flow rate of the bypass to infinity (ha ha) and temperature depression resulting from dilution from the rad return tends to zero. What you sought to achieve, unless I'm mistaken, is a warmer radiator return (thanks to the position of the thermostat) and effectively a greater bypass flow (again due to thermostat position - and the additional bypass pipe linking the radiator top and bottom hoses). Net result is that coolant at the original thermostat location, where the original bypass blends with the radiator return, is a good deal warmer than it once was, which effectively moves the cyclic minimum of the temperature oscillation upwards, diminishing the temperature oscillation amplitude. My interpretation is that this is a system that is much better damped than standard. Regarding efficiacy under extremis - I was picking up on what Will was saying - but I very much doubt that your additional bypass hose will compromise cooling ability to any great extent (at least not in Northern European driving conditions!) |
| Rob Bell |
| Dave why were MGR such a failure at solving this problem. Secondly this has been a problem with K series engines for all time even before they were fitted to Fs so why do engines fail in "normal" cars. |
| Bob (robert) I am coming out!! yes I once owned an MGB!! |
| <<why were MGR such a failure<< haha. They individual engineers probably didn't have the motivation to avoid HGF like we do! I don't know. Failure on 'normal' cars is for the exact same reason - the thermostat is on the wrong side of the engine still. The effect is not amplified by the large volume of cold coolant that the rear engined cars have, so is less frequent. Rob, The whole story can change if we start extrapolating to infinity... ;) Within the confines of reality, it is more practicle to move the thermostat than try and increase the bypass tube flow rate to counteract the cold inrush. I wasn't aiming to increase the bypass flow - I was aiming to reduce the volume of cold water available in the system 'behind' the thermostat - the position of the thermostat does this. The bypass here has an easier job of bleeding hot coolant (of which we now have plenty) into the cold coolant remaining. Anythat doesn't mix straight away has several feet of hot coolant to fight with before it reaches the block. Im not sure increasing your bypass in its original position can create an indefinite supply of homogeneously mixed cold / hot coolant - remember just how much cold coolant is still in the system. Dave |
| Dave |
| I wonder if all of the termal effects you have shown with your remote thermostat can be obtained -much cheaper- with the original thermostat drilled. I ask that question because after the HGF on my 96 F I had the thermo drilled and since then and 60k after have had no further problems. Heavy foot and frequently trashing from cold to be considered too. |
| Andrew Bowles |
| OK Dave you are convincing, now if the block and head were made of steel would the "thermal shock" cause the same problems? |
| Bob (robert) I am coming out!! yes I once owned an MGB!! |
| Rob: Laminova kit being "parked" for the moment. I think the front mounted PRT will give a bigger gain in reducing thermal change on the block/head than a dual oil/water warm up. Long oil lines are not a very elegant option. I'll keep it and see what happens, unless anyone wants it. Bypass flow can't increase, so not worth thinking about? As to total cooling capacity, nothing has changed, and I don't think the original rad is inadequate for normal fast road use. David/Rob: my point about using the TF PRT, but run backwards, is that at max water temperature the bypass will shut, all flow will be through the rad at full bore. PS my car is slowly losing water, no obvious leaks anywhere! No emulsion either. Can water disappear internally? Driving around with the lid off, looking at every opportunity for leaks. Damn this car, I have a PRT in my hand and it sounds like I may be a suitable canditate for HGF. Views anyone? |
| Jerry Herbert |
| >>max water temperature the bypass will shut, all flow will be through the rad at full bore<< Even with the small bore bypass still open, most flow will be through the big bore - so much so at operating temp the short section bypass I have used is soft. You could fully close it with a normal 'valve' type thermostat (eg 620Ti) - im uneasy about commenting on the PRT, because I don't know how it works well enough to feel I can. >>ow if the block and head were made of steel would the "thermal shock" cause the same problems<< The shock would still be experienced, but the effect may not be twisting or creeping as with aluminium... bit acedemic really... I've never heard of a iron engine with the thermostat in the wrong place though! There are plent y of Alu engines out there that are fine though. Dave |
| Dave |
| >> Rob, The whole story can change if we start extrapolating to infinity... ;) Within the confines of reality, it is more practicle to move the thermostat than try and increase the bypass tube flow rate to counteract the cold inrush. << LOL quite true - and I see where you're coming from - you've got a practical, low-cost solution, which is what is wanted at the end of the day. :o) It will be very interesting to see how the PRT performs... |
| Rob Bell |
| >>As to total cooling capacity, nothing has changed, and I don't think the original rad is inadequate for normal fast road use.<< Ah, but something has changed - the resistance to flow of the new 'in-line' thermostat location is a fraction of that from the old location. Cavitation effects caused by poor flow into the pump is also eradicated - ie the pump is more efficient, the flow is more laminar, and hence the cooling capacity is increased. This can be seen in the graphs on my site. The knock on effect is further flexibility in the system - ie sudden increases in temp can be absorbed more easily, and the systems ability to cope with very hard use is improved. Dave |
| Dave |
| Hi all, have with great interest read thru the comments to Davids excellent work. It IS a worthwhile improvement to the cooling system that probably will reduce or even eliminate the chance of a HGF if otherwise normal service and care is used. Did this alteration to my system 5 years ago,(see Dieters site)with inserted temp-sensors and logging. It all started with temp-sensors only and a session at Knutstorp raceway here in Sweden. What I saw was alarming,the thermal shock to the system after going down the start/finish-straight at full speed was just not healthy for any engine ,being alu or iron or a mixture of both!! So during the winter break I started to investigate if there could be done any "easy" alterations to the system. I went down a slightly different road than Dave, keept the original thermostat BUT fairly large holed (4,2mm)and changed for 82 deg.C . On the market there are several suitable external thermostats,mainly for older cars. The Rover SD1 as well as LADA will be OK. Especially the Russian LADA item was interesting as it was very well built,had low temperature and did not close totally but had a "jiggle valve". Making a long story short , it works as intended ! Maximum temp cycling when at temperature at hard track driving is +/-3 deg.C ! Of course this small temp cycling will also be benefitial in everyday driving,so far no trouble (touch wood!) on a soon 10-year old VVC. |
| Carl |
| Yes - the 'remote' thermostat from most of the Rover V8 engined cars - range rovers, SD1's etc... Lada is a good source, 80-90's BMW's, and various others... I prefered to make my own housing, mainly for cost reasons, but I wanted to be able to experiment with different thermostats if required - the assemblies used on other vehicles 'include' the thermostat - ie it would start to get really expensive, really quickly! Dave |
| David Monks |
| I know that was asked before but with no answer. How would this remote thermostat compare to a drilled thermo? |
| Freddy Christmann |
| Does anyone have a part number for a remote thermostat that can be fitted to the MGF with as little trouble as possible? There has been a part number given for what was called a PRT (Pressure Relief thermostat) in the posts but I thought this was a different type of thermostat. From David Monks last post: "Yes - the 'remote' thermostat from most of the Rover V8 engined cars - range rovers, SD1's etc... Lada is a good source, 80-90's BMW's, and various others... " |
| KS Gould |
| Not that easy to answer your question Freddy,but something like this: The drilled thermostat obviously avoids total stop in the flow when drilled with approx. 3-4 mm hole. Gives slightly longer warm up and combined with a never fully shut heater knob smoothens out the system. But much of the "cold water shock" as well as not perfectly maintained "servo loop" for the whole coolant system makes it not the best solution. The thermostat is still in the wrong position... A remote thermostat is probably at the moment the best. ( I did run my car with a drilled 82deg. one season before the external thermo work) BR./ Carl. |
| Carl |
This thread was discussed between 09/03/2006 and 16/03/2006
MG MGF Technical index
This thread is from the archives. Join the live MG MGF Technical BBS now