Exhaust manifold over heating?

Viewing 20 posts - 1 through 20 (of 20 total)
  • Author
  • #392723

    After running about 20 miles with our 1916 48B, returning to the garage to check the engine temperature with a laser thermometer.

    The radiator is about 150degrees at the lower outlet. 170degrees at the top neck. Engine is between 170-200degrees. Seems ok?

    However the exhaust manifold at the rear cylinder is at 450degrees! Could the be normal? Oe am I running lean? What should I do?

    Never been exact where the rich/lean on the steering column adjustment? The car starts super easy. It seems there should be a setting to start and another when running.


    Exhaust manifold temp over 450 is fairly low, I have measured my ’35 845 at over 600F standing still. It will get hotter than that when being driven on the road under load. The exhaust gases flowing fast through the inside of the manifold will average on the order of 1000F, the manifold can only be cooled below that temperature by the slow 140 degree air coming out of the radiator wafting over the outside of the manifold and by conduction into the 200 degree block.

    Note that high performance engines running at high power will make the exhaust manifold glow cherry red – ~1200F, but the block will be kept in the 200’s by the cooling.

    The reason exhaust manifolds are susceptible to cracking is because of the large temperature difference between the manifold and the block they are attached too.

    The manifold sits at the same temperature as the block until the engine is run, then quickly gets a lot hotter and tries to grow from thermal expansion which the connection to the cooler block prevents. This puts a strain on the manifold that may eventually crack it.

    It looks like you have a good 20 degree temperature drop through the radiator and the block is being cooled down normally to a 200 degree temp via the flow of water starting at 150 and regaining the 20 degrees when it goes back to the radiator.



    Jim, thank you for the answer. I had not checked the Pierce exhaust manifold and noticed the “Heavy-Lean” adjustment on the steering column. I had it set closer to the “Lean” setting and not changing it for starting and running.

    I should not be concerned about overheating.

    I think, and am asking, where should the “Heavy-Lean” be set when starting? When when running?


    Bill, I don’t have any experience with these older cars, and hope one of the members with experience on them will weigh in.

    Presumably the mixture should be set to rich (heavy) for starting and warming up and leaned out for driving.

    Speaking just in general terms for most any engine, a lean mixture will improve fuel economy for cruising down the road, will burn cleaner leaving fewer deposits and less spark plug fouling, but won’t develop as much power for climbing hills at wide open throttle and as you know will increase exhaust temperatures, be more susceptible to exhaust valve recession and knock (ping). However, I don’t think these early engines should be at any risk of that using current 87 octane fuel.



    I agree with Jim on the fuel. Tetraethyl lead wasn’t in gas until the 20’s so when the care was new it ran on lead free gas. 1922 is the first usage of lead I have read.


    Run you early engine lean at huge risk…………melting pistons, burning valves, and cracking castings is common. Octane today is made using alcohol…….which runs hotter and leaner than pure unleaded gas of the old days. The car should ALWAYS be set to run rich. Modern fuel had different specific gravity and most importantly different heat and power content with modern blending stocks. I could go on……….to properly set up the car, a five gas analyzer is best if you can gain access to one. Running some upper end lubrication is not necessary, but on very early engines I add it for safety’s sake. It’s a hundred year old car, treat it with respected don’t pound it down the road, or very unpleasant things WILL occur.


    With my 1915 Model C-3 I have always run the car with a mixture of regular gasoline and Pennzoil two cycle oil in a mixture of 100 to one so there is a small amount of oil in the gasoline.The Series 3’s and earlier do not have the mixture control on the steering column as do the Series 4’s.I would find a source for straight regular gas with no ethanol.Some more rural gas stations may have it for tractors and farm machinery.Gasoline in the early days was more oily.My car has a nice tan deposit on the exhaust pipe and a friend remarked my car seems to run very clean.I have the belly pan which helps to hold heat in and the carburetor is heated by the car’s cooling system all making for good combustion.Too much oil in the gas and your Pierce Arrow will smoke like a Trabant.


    Early cars were built with loose tolerances. Upper cylinder lubrication

    happened by oil that got by the rings. To go really fast you built an engine

    loose as a goose (witness the starting line at early races and the volumes

    of exhaust fumes).Scraping carbon build-up was common maintenance. About 1920

    Marvel Mystery Oil made it’s appearance. Today we build engines with lighter

    pistons with modern rings at tighter tolerances that rev higher and prevent

    oil from passing through to the top end of the engine. I don’t leave the gas

    station without adding Marvel Mystery Oil and 100 years later it’s readily

    available even if you want it by the gallon.


    Early motorists bought their gasoline in 5 gallon tins from drug stores or

    fuel distributers that the city fathers made sure was far enough from town

    to avoid a conflagration. Fuel quality changed from location to location.

    Motorists carried a chamois to strain out impurities. The early gas was about

    50 octane. It would compare with white gas used in Coleman camp stoves. It

    burned very rapidly and suited the need for the flame to travel a long

    distance in a T head engine. In the twenties, compression went up and the

    fuel caused pre-ignition. Lead was added and octanes (slower burning fuel)

    went higher. With hundreds of thousands of cars on the road, Service stations

    were introduced in 1914.


    I agree with Tony and Bill, and like Bill I use two stroke oil…………


    There was a spark plug on the market (“Colortune”) which had a clear body, for tuning. You could actually see the flame color and adjust the mixture at idle, but it would not stand much pressure or heat so it was for tuning at idle only. I have one somewhere but I have never tried it; wonder if anyone has?



    Leaded gas is available through small private airports, but they may not pump it into automobiles.

    You may need to bring a few 5-gallon jugs.

    Below is a website for Pure Gas!




    Lead is only needed on pre-war engines to prevent exhaust valve seat recession, and the contemporaneous engineering literature indicates this did not become an issue until the 1930’s. It is caused by a combination of high exhaust gas temperatures, RPM and lean air fuel ratios. Even though the lower compression ratio’s of 1910’s era cars start out theoretically to have higher exhaust temperature than 1930’s, the lower RPM and high surface area’s of the combustion chamber quenches the mixture more by the time the exhaust valve opens. Another factor was likely the relatively poor fuel vaporization leading to less complete combustion. Liquid fuel in a rich mixture that doesn’t burn completely will quench temperatures and combustion. It also reduces the free oxygen that induces valve seat recession. Valve recession should not be an issue for 1920’s and earlier cars, and should not need special efforts to get 100LL aviation fuel or other additives.

    The original purpose of lead in fuel had nothing to do with valve seat recession, it was to raise the octane and prevent knock, allowing higher compression ratios. Its benefits to valve seat life was not understood before WWII and played no role in whether an engine was “designed” to run on unleaded. As Tony points out, regular gasoline in 1925 only had an average octane rating of 55, premium was 71. It had risen to 72 and 77 respectively by 1935. An antique engine running 87 octane now has a serious problem if it exhibits any knock (ping). One of the problems can be buildups of carbon deposits in the combustion chamber from running too rich or oil blowby (or adding extra oil to the gasoline). The deposits can glow and cause pre-ignition. Early piston rings had poor oil control and typically were “de-carbonized” by removing the head at 15000 miles or so. If you have an engine that “Diesels” – keeps trying to run after ignition is cut – it is a sign of excess deposits.

    Another factor is spark timing. A retarded spark increases exhaust gas temperature. This caused lots of heat problems in California in the ‘70’s and ‘80’s when many 60’s to 70’s cars were forced to have spark retarding devices to reduce hydrocarbon emissions. An antique in fine fettle running on 87 octane that was designed to run on 55 or 70 octane should be well clear of any knock issues and the spark can be advanced 5 degrees or so from original factory.

    I think alcohol’s and modern fuel blends influence is mis-understood. There actually is very little difference in the energy content of the wide range of hydrocarbon constituents used to blend gasolines then or now. With perfect stoichiometric combustion of a vaporized mixture (not lean, not rich) ethanol actually has a tiny fraction (.1%) less energy per volume of mixture than gasoline. At the heavy end kerosene is about 2% less than gasoline, and at the lighter end butane is 1% less than gasoline. Butane is typical of the lightest (volatile) end of the many constituents of gasoline that evaporates very easily at low temperatures and is needed to start a cold engine. Ethanol does lean out a carburetor jetted for a rich mixture and run closer to stoichiometric, so it makes sense to adjust the mixture a bit richer, but remember the ethanol is only 10% or less to start with. A well controlled carbureted engine test done in 1941 showed a 10% ethanol blend increased net power by about 3 ½% after increasing the compression ratio slightly from 6.0 to 6.2 to obtain the same knock limit.

    I think the main difference is not the basic energy content of the fuel but its volatility – its ability to atomize and evaporate in the carburetor and manifold to form a combustible mixture. Modern fuels in general tend to be a bit more volatile and evaporate/atomize more completely and this can increase power and temperature a bit (and creates the vapor lock problems). This is true of both conventional gasoline and ethanol. If the fuel doesn’t have ethanol it still must have hydrocarbon constituents in the same general range of boiling temperature as the 10% ethanol. Historical data shows that fuel volatility increased steadily from 1946 to 1965, well before ethanol was a factor.

    I’m not defending ethanol, just saying it seems to be the go-to boogie man for all manner of engine problems rightly or wrongly.



    Energy content can be broken down into energy and heat content, and the E10 has more heat and less energy. Blending stocks, mole weight, and fuel density are all also different. Lead was NEVER added to fuel until 1937, and then in very small quantities in limited areas. Old style piston ring packages, distributor curves, and timing were also affecting performance. I have ran Pierce Arrow cars on my chassis dyne, and have much different real world figures than the engineering back of the envelope fuel calculations. We compared five different fuels, all with different blending stocks. At the rear wheel figure 7 to 9 percent less power with E10. John and I have spent years configuring fuel, ignition, and engine components to get maximum performance and reliability results. Perfect stichometric efficiency in a F head is simply not possible, and the ratio is in the 12’s to low 13’s. All done with a five gas analyzer and chassis dyne.


    PS- I mostly agree with what Jim posted.


    For what it may be worth: Bill, on my 1918 dual valve I run rich (“heavy”) to just below the point where there is visible exhaust. To see if you’re running too lean (“light”), with a warmed up engine running in 4th gear at 30 mph, suddenly floor the accelerator–if it backfires, it’s too lean, and if it stumbles/loads up, you’re too rich.

    The higher your speed, the more you should enrichen the mixture. These knobs do NOT function like choke knobs, which you can forget about once the engine is warm; be thinking about how rich/lean you’re running and tweak the knob every so often. Those adjustable-main-jet carbs are very handy at altitude: On the Modoc going thru 6,000 ft and on Glidden in Idaho last year at 7,000 ft, I could lean out during climbs and richen as we descended. Fixed-jet cars were puking black smoke at 7,000 ft.

    There’s a near-infinite range of adjustment due to the clamp on the main jet rod. Supporting Ed’s idea of a gas analyzer, it might be useful to have that done ONCE and change the clamp position so that optimum adjustment at perhaps 40 mph UNDER LOAD at your home altitude represents “40%” rich on the amount of travel of your heavy-light knob. I do that on Series 80 rich-lean levers by ear and vacuum gauge, as I don’t have a gas analyzer. At least you’re in the ballpark.

    Better to pump out some black smoke occasionally than to burn valves–and in dual valve engines, a too-lean mixture will often lead to a cracked block.


    In between the lines, Ed has touched on the fallacy of the ethanol addition to gasoline. Less efficient, so in the end poorer gas mileage, at the expense of the corn market and the subsequent land clearing, fertilizer plant pollution to support additional corn harvest, and so forth.

    Luckily, we have local stations with 100% gasoline, including a Walmart service station. I run it on the old cars all the time…


    I hope that Bill has gotten the basic answers to his original question before going further down the rabbit hole.

    Ed, I am curious how you come to the conclusion that “Energy content can be broken down into energy and heat content, and the E10 has more heat and less energy.” Analysis of the thermodynamic properties of the combustion energy and products of combustion do not support this. Was this from observing the reduction in maximum power combined with an increase in measured exhaust gas temperature? If so, I assume the EGT was measured at the tailpipe with the gas analysis probe?

    Just trying to untangle this to better understand what is going on.

    The table shows the basic thermodynamic properties of the combustion products of pure ethanol and gasoline under perfect conditions for a fixed volume of mixture. The fundamental properties -specific heat coefficients, ratio of specific heat coefficients, and temperature rise with combustion – are remarkably close to each other.

    This tells me that it is not the basic properties of combustion products that are causing the problems but the myriad other things going on in the engine that of course deviate from the ideal.

    I suspect that the problems stem more from the differences in the blends at the carburetor and intake manifold before combustion. A reduction in power with an increase in exhaust gas temperature is probably related to slower and/or incomplete combustion coming from the fuel blend changing the atomization of the fuel in the intake manifold and messing up the fuel/air distribution to each cylinder. The higher heat of vaporization of ethanol dropping the temperature of the mixture while passing through the intake manifold is probably one factor. Ethanol does have a higher autoignition temperature that could come into play. Slower combustion from poor mixing can reduce power with an increase in exhaust gas temperature and incomplete combustion can keep burning in the exhaust manifold after leaving the cylinder – increasing temperature without adding power.

    Perhaps this seems moot, but if we need to operate on current fuels it helps to try and understand what is going on.



    Hi Jim,

    One item missing in the table is ‘how’ or ‘in what type of chamber’ the gasoline mixture and the ethanol mixture were burnt or tested. ?

    I’m sure somewhere on another related page from whatever document or study you got the table from, will have the info and hopefully describe the testing procedure and methods.

    The main thing I see is that the air to fuel ratio says VOLUMES about the amount of ethanol it takes to make a powerful, burnable mixture.

    For gasoline a ‘perfect’ mixture [Stoichiometric] is shown as 14.88 parts air to each part gasoline.

    But for Ethanol, a Stoichiometric mixture is roughly a third less air for the ideal fuel ratio: 8.95.

    What this means to me is that what i have always thought about ethanol as a fuel rings true: it takes almost twice the fuel to create the same power. An engine breaths AIR, the amount of fuel needed to combust the air + fuel mixture is irrelevant to the engine, it is just an air processing machine. more air in with a burnable mixture of fuel will equal more power to the crankshaft, This is why a 400 cubic inch displacement engine with everything else equal, will make twice the power as a 200 cubic inch engine.. It’s all about processing the air/fuel mixture.

    So, to get the same power out of an ethanol burning XXX cubic inch engine, to match the power from burning gasoline in the same displacement engine, will need almost double the ethanol vs gasoline.

    I know, it’s not ‘double’ more like ? 80% more, I hate math. But suffice to say the amount of ethanol for the volume of burnable mixture is considerably more than pure gasoline.

    Jim: you did mention time to burn, rate of burn, higher temperature to ignite etc for ethanol over gasoline, and these are the true problems with ethanol in our pre-WWII engines.

    The main thing is compression ratio. Alcohol needs 14:1 and higher to burn well in a combustion chamber in an Otto-cycle engine.

    Our Pierce Arrow engines had 4:1 up to 6.4:1 compression ratios. Ethanol just will not burn well, completely or effdecenty with such a low compression ratio.

    I think that what you mentioned, and I have commented on before is the likelihood of still burning ethanol being exhausted into the exhaust manifold with resulting much higher exhaust manifold temperatures and therefore much higher under-hood temperatures.

    I agree that for most states, we are stuck with ethanol-tainted gasoline. To me this is criminal for us to be forced to support the ‘ethanol’ fuel federal government program with our tax dollars and also to not be able to buy ethanol free fuel for our chain saws, generators, lawnmowers, leaf-blowers, boat motors, snowmobiles, off-road motorcycles. NONE of the ‘other’ devices that burn gasoline run well, or even survive on ethanol-tainted fuel..

    Every gallon of gasoline that is mixed with ethanol is subsidized by our tax dollars at $.40/gallon.

    I’ll be interested to find out how the fuels were tested to create the numbers on the chart !!

    Greg Long



    To try and untangle what is going on first we need to understand the basic properties of the fuel and air under ideal conditions. Then we can use that starting point to try and sort out the huge number of variables that happens in a single cylinder of an actual engine, then we compound the problem with multiple cylinders, carburetor and intake manifold.

    I get the impression that general observations and tests of a few specific engines are being used to try and understand the fundamental properties of the fuels, rather than first understanding the properties of the fuels and then use that knowledge to try and understand what is happening in the specific engines. A famous example of the difference is James Watt. Steam engines had been pumping the mines of Britain for 60 years when Watt used the fundamental properties of water’s latent and sensible heat measured by he and Boulton to understand why Newcomen engines were wasting 95% of the heat energy available from the coal to heat the water and generate steam. He used this understanding to invent the principle of external condensing and doubled fuel efficiency -wasting only 90% of the coal’s available heat energy. It sparked the industrial revolution.

    For 40 years as part of my career, I did thermodynamic modeling of engines, primarily aircraft gas turbines (fanjets, turbojets and turboprops) but also piston engines. This to predict the thrust and fuel burn for an engine flying at 55,000 feet Mach 1.5 (for example) from data originating from an engine run motionless in a test cell on the ground. Sounds complicated but gas turbines are not nearly as complicated as antique piston engines! My first foray into engine research was trying to do a computer model of the products of combustion in a single cylinder piston engine and then testing the engine on a dyno. The model of combustion had to iterate the equations for 14 separate chemical reactions with 14 unknowns, and shall we say, taxed the IBM 360 mainframe I was using.

    The combustion energy comes from what is known as a bomb calorimeter. It is a fixed volume container of fuel and excess air mixture placed in a water bath and the rise in the bath temperature is used to measure heat of combustion. It is the standard method for measurement of basic energy from combustion. That may seem like a very esoteric method that doesn’t relate to what happens in a real engine, but how can one derive the characteristics of a fuel by looking at dynamometer data from a specific multi-cylinder piston engine with unknown mixture distribution, big cooling losses, internal engine friction, etc etc? Your business aircraft engine performance was based on assuming a standard 18,400 BTU/lb for jet fuel weighing 6.7 lbs/gallon derived from these standardized tests. A remarkable thing about hydrocarbon fuels is how little difference there is in basic heat of combustion per pound from the lightest methane to bunker “C” crude used in Diesels.

    The flame temperature rise I put in the table is theoretical as it has the simplifying assumption that the piston comes to a complete stop while the perfect mixture is burned , there is no heat transfer, the combustion is perfect with every molecule of air finding a fuel molecule, and ignoring some extreme temperature effects. For simply comparing the actual fuel properties to one another, these factors are niggling.

    As you point out the piston engine is basically an air pump. Each stroke it tries to pull in a fixed volume of fuel-air mixture to trap. Of course, there are a lot of factors that come into play on just how much air it does trap on each stroke, but for our purposes I think a comparison of the energy in the air-fuel mixture based on identical volume is the only rational basis. That is why I used the energy of a fixed volume to compare.

    In the table, note that even though the ideal air/fuel mixture is 8.95 versus 14.88 in the table, the mass of air in the volume is actually only 5% less for the ethanol than gasoline. 58% more weight of ethanol was put in the mixture which displaced only 5% of the air because there is a lot more air than fuel. On a weight basis the ethanol only has 63% as much combustion energy as the gasoline, but 58% more ethanol was added, and the chemical balance of the equation works because the ethanol carries an extra oxygen molecule with it. This is for a mixture already vaporized. Basically in a carburetor it would mean jetting for a richer mixture but the engine is still pumping the same volume.

    In a real engine, ethanol’s heat of vaporization is higher, cooling the mixture more as it vaporizes from liquid, theoretically dropping the temperature in the manifold and increasing density and the mass of mixture drawn in with each stroke. This would increase power a bit, if it happened that way. I have a 1920 SAE report on what it took to get the best power and fuel economy using pure ethanol in a tractor engine, and basically they had trouble vaporizing the ethanol and had to add more manifold heat. They also had to increase compression ratio (as you noted) due to the higher ignition temperature. They raised it to around 5/1 (stated as a compression pressure rather than ratio). In the end they got very good results with comparable efficiency after a lot of cut and try experiments. It included adding transparent windows in the intake manifold to observe un-vaporized liquid fuel running down the branches of the manifold. That was on pure ethanol. With only 10% ethanol in a gasoline blend that still has the more volatile gasoline components to light things off initially, I don’t know how much of a factor higher compression ratio and the higher ignition temperature is.

    I can understand that 1920’s and earlier engines could have a tougher time on ethanol blends. With the very large combustion chamber and little turbulence the mixture is not going to finish mixing very well, having rich and lean spots, thus to insure combustion every time without misfire needs a rich mixture to make sure an ignitable mixture is at the spark plug. When it does burn it may burn slower and less complete due to the non-uniformity, even though ethanol itself has a higher flame speed. The mixture distribution to each cylinder from the manifold isn’t very uniform to start with compounding the problem. Besides higher compression ratio, a later 1930’s engine with Ricardo high turbulence “squish” chamber at least will better mix up the fuel and air near top dead center just before and during combustion. The ethanol may screw up the mixture distribution at least in part due to it cooling it a bit more and inhibiting vaporization. Ironically it might take more manifold heat to improve the situation, giving more vapor lock problems. There are potentially some other solutions, but they would require cut and try.

    Even when running richer than stoichiometric a real engine will not consume 100% of the available oxygen and will leave some in the exhaust that may burn while traveling out to the tailpipe, much as air pumps were added in the ‘70’s to finish burning rich mixtures.

    It would take a pretty extensive test cell and instrumentation setup to run these things to ground on any given engine, which is probably not in the cards.


Viewing 20 posts - 1 through 20 (of 20 total)
  • You must be logged in to reply to this topic.