Approximately the first 2500 Model T engines were designed to use a water pump. I suspect, but I do not have any documentation to support, that one of the major reasons Ford discontinued the water pump engine was to save money (fewer parts needed for thermo-syphon engine cooling). There was the initial cost of redesigning the engine block and head but that would have been recouped easily over the millions of T's that followed.
Does anyone have any documentation, period articles, practical experience etc. that would help me better understand why he discontinued the water pumps? Is my guess about saving money part of why he did it? Did the water pumps prove ineffective?
I have read in recent articles that the water pump engines tended to sometimes over heat. But from a single sentence in an article it is difficult to know any details. I.e. if the owner let the water leak out of the system -- it clearly would overheat. But did it over heat when functioning properly? [And I believe but have not tested that even the thermo-syphon system will tend to boil / over heat on a hot day climbing at high altitude in the Rockies].
Any thoughts and especially experience or early documentation would be greatly appreciated.
No, I'm not thinking of purchasing a water pump T (I wish...) I'm trying to make sure that I don't repeat "legend" that isn't true. [Remember all the bad press about the Model K that turned out to not be true?]
Hap l9l5 cut off
No documentation on my part but besides cost savings having no water pump I believe Ford intended the Model T to be as simplistic as possible so the average farmer or knowledgable person could fix or repair most common problems with minimal hand tools just about anywhere. The less parts, the less to break and the less to repair.
I think it's because he saw all the anti-waterpump posts on the forum and decided that they'd best be used as doorstops in his factory instead of on his cars
But really, this is something I've always wondered, but never thought to ask. I hope someone has a good answer!
My guess would be, with the seal technology of the time, most were rope seals, the pump may have had leak problems. Even the Model A pump leaked with the original seal design. It was easier to omit the pump.
With the low-compression and low-revving T engine, there really isn't that much heat generated when compared to a modern engine. I think that Ford just decided that the pump was unnecessary and discontinued it. I have driven several different Ts, and have come to the conclusion that a T will not overheat under any normal situation if the radiator and the rest of the cooling system are clean and in good condition. I'm not anti-water pump, but I feel that it's a patch on a cooling system with other problems.
Not to be forgotten, there were a number of engine manufactures as well as automobile manufactures who used the thermo-syphon cooling system.
And it should not be forgotten that during the time of the introduction of the Model T there were a large number of now forgotten automobiles that used water pumps.
Is it also possible that as more cars were produces and driven in cities and large towns traffic jams and stop and go traffic lead to the addition of water pumps beginning in the 1920's as an accessory?
Its incredible to me how well a stock T works but it was built to move faster then a horse running out of energy on roads. With more folks moving to towns off farms there was not much room for animals Ford hit the market at the right time. I agree with Michaels post.
I have no documentation, but I would agree it was probably in the interest of eliminating needle$$ Ęo$t$.
Very smart topic ... seaching for the "official reason" ! I guess that the only official answer could come from Ford Motor c.?
Of all the engines I have had experience with, only my T and John Deere A have I not had any problems with a water pump. I think the only disadvantage to a thermo-syphon was the need for the radiator top tank to be rather higher than the water jacket on the engine. As far as why Ford eliminated the water pump is simple, it eliminated the cost of a machined casting, two bearings, a shaft, water seal, pulley and impeller.
Does the water pump rob horsepower? I would want all the power I can get when you have only 20.
Attached a reference to an article in 1908 publication "The Autocar." pp 284 - 286.
There is a reference to the use of thermo-syphon and the Perkin radiator tubes.
The best part of the 1908 article talks about the steps to improve the thermo-syphen system compared to the "pump system."
The magazine had other articles praising the virtue of the thermo-syphon system. I am certain that Henry and the boys were reading such publications for ideas on improving their designs.
https://books.google.com/books?id=4n8fAQAAMAAJ&pg=PA284&dq=thermosyphon+system+1 908&hl=en&sa=X&ved=0ahUKEwjfgpnGyp3XAhXp5oMKHW5aAlwQ6AEIJjAA#v=onepage&q=thermos yphon%20system%201908&f=false
Hap, I've never heard of or seem any "official " reason. As we all know, Ford was always interested in cost savings, and an improved product. The pump is a 5 piece casting plus the shaft and extra gear, so there was substantial cost savings and a cooler running engine! A win win for Ford.
It has been my belief that Ford wanted to make the car economical and he could save a few dollars by not installing a water pump. It is a well known fact that a T in good tune and without leaks and with a clean radiator will run just fine and not overheat without any water pump.
The highest altitude I have driven a T without a water pump was to Cedar Breaks in Utah. It was about 10,000' elevation all uphill. The T did not overheat.
We also have many hills to climb in California not quite that high altitude, but sometimes very hot weather and the T doesn't overheat without a water pump.
Having said the above, I run 50-50 Prestone anti-freeze in my cooling system which is known to raise the boiling point of water. Another fact is that water boils at a lower temperature at high altitude, so the cars running only water might boil at higher altitude even though they are not overheating. The modern car has corrected for that by using a pressurized system which also raises the boiling point of water.
Interesting to me that the word Perkin is in your post George. My 151 Perkins Diesel has 17 to 1 compression with a combination of thermal Syphon and water pump system. It never over heats wonder if there is a connection or not to the English engine maker?
While many might discount the above mentioned theory of water pumps being used as "door stops at the factory", one must realize that Piquette has those heavy sliding fire doors. I'm guessing one of those doors slammed shut on the side of Model T #2500 right in front of Henry Ford.
In the online encyclopedia, the notes for 1908 say:
The following was supplied by Trent Boggess and is excerpted from Frank Hadas Reminisces at the Ford Archives. Mr. Hadas was a chief tester during the NRS & K period at the Piquette Avenue plant and also tested the first Model T's. He writes:
"On the Model T, we had trouble with the fiber discs (bands?) on the planetary transmission. We had trouble with the damn pump and a lot of things."
The encyclopedia also mentions that on January 19, 1909 the design for the thermo-siphon engine was finalized. Being so hot on the heels of the introduction of the car I wonder if that wasn't the plan from day one but teething issues forced production to start with water pumps and they would fix it in production when they could. Is the water pump from an early T the same unit as found on a NRS?
In speaking with the owner of an early car that gets driven, I think it's A + B + C = no pump on the later cars:
A: additional cost that Ford's supplier(s) may have increased after the initial run of 2500 was used up. Or, possibly, the supplier(s) were unable to fill a subsequent larger order. This was a problem that worried Ford throughout Model T production, and was a major reason that Ford insisted on multiple suppliers for a single part or assembly.
B: Ford's desire to simplify, simplify, simplify.
C: The pump was inadequate for the job. The impeller design was inefficient, besides blocking the flow of water after the engine was/is turned off, resulting in hot spots. This may be the reason for so many cracked early heads. The owner told me that merely stopping at a red light means instant overheating.
A letter was found in the back of a ledger from the Ford Company dated 1908 1 April.
" As I write this for future historians I want to say history is bunk. To offer proof, at the end of production number 2500, I will be introducing a system for cooling the motor called "Thermo-syphon." I am removing the water pump, this should create quite a discussion for the future whippersnappers." Henry Ford.
Interesting topic for research. All Fords before the Model T had water pumps. I believe the design of the pumps was basically the same three bladed impeller, but driven by different methods. The Model T water pump was part of the cylinder casting and driven directly by the timing gear. The rumors say Ford made a lot of running changes on the first Model Ts. One presumes this change was done for a reason, like going from two pedals and two levers to three pedals and one lever. Now, if we could just find out what that reason was.
The "fiber discs" are the clutch discs between the clutch plates on the NRS cars.
(Message edited by paulmikeska on November 01, 2017)
Maybe Ford did some tests without a water pump and it worked just as well without one. And then he asked himself why does it need it. Evidently it did!
I meant to say that it didn't. Sorry about that.
I had them on my 2 black era T's. After I put a new radiator on one and a recored radiator on the other without the water pumps my cooling issues were over.
A lot of good speculation above, a few very enjoyable snide remarks (I like snide remarks, as long as they can be taken for the humor intended) (George J D, I like the April 1 '08 date of your "letter"). So, let me add a speculation on something I noticed several years ago.
For whatever reason, I see almost everything as engineering, laws of physics, and math. Just my unexplainable nature.
Hap Tucker in his first paragraph mentioned "There was the initial cost of redesigning the engine block and head but that would have been recouped easily over the millions of T's that followed." He of course was referring to replacing the first early engines with a better design for the remainder of production. That redesign was virtually unavoidable as a big mistake had been made in the rush to design and build the new model T (some "rush", that initial design took about two years as it was). The previous engines (A/B/C/F/K/N/R/S) had all been cylinders cast singly or in pairs. This new engine was four cylinders cast en-block. The smaller block pairs (and singles) with individual water inlets and outlets flowed water reasonably well and cooled okay (Mostly? There were some issues in some early models). The N/R/S models, similar to the model T in general configuration, had separate water pathways through each pair of cylinders. In the rush to production of the model T, with all the many changes being made, something was missed in the water pathways. The water pump pushed cooler water in at the extreme front of the block, pushed through the block, into the head, and out the front top of the head. Even with some restriction between the block and head at the front, the water tends to take the shortest and least restrictive path to the radiator. The front half of the engine with the inlet and outlet connections gets cooled very well. The rear two cylinders do not get cooled reliably. Engine speed becomes critical to a consistent flow as changing speeds causes the flow to stall in some areas of the engine and not others. The engine running at constant speed can also result in the flow stalling in some areas of the block and head. Thermo-syphon will tend to kick in within the block and mix hotter and less hot waters before it exits to the radiator, but that will only mitigate the over-heating a little bit.
The rest of T production, dumped all the water in the middle bottom of the block, with some restriction in the middle between the block and head making the water move throughout the engine and head. Additionally, the thermo-syphon tends to favor moving the hotter water the most, allowing cooler areas to bring their temperatures up to a closer average. Water pumps tend to move water based more upon circulation forces and restrictions, cooling some areas more, and other areas less.
I have dealt with these issues before, several times, on both antique and modern car engines. Many years ago, I had a big Chrysler wagon as a family car. We took a 300 mile drive up to see family, on the hottest day of the year. Naturally, being the hottest day of the year, the thermostat decided to fail (and of course, naturally, it failed closed). The car never actually boiled, until we arrived at our destination city and slowed down from highway speeds. Instant full boil. After some diagnostics, discovered that the design of the (failed) thermostat not only closed off the main flow of coolant, when closed (failed or otherwise), it opened a "bypass" to hopefully slowly warm the coolant in order to open the thermostat. The result of that design was that the bypass cooled the front half of the engine well enough, but the back half hardly cooled at all. After a couple hundred miles on the hottest day of the year? The engine was toast. Why had the idiot light not signaled a problem? Because the idiot that designed it put the sensor right where the bypass showed coolant temperature was just fine instead of where the coolant hardly circulated at all.
Many antique mono-block engines have multiple port either inputs or outputs, with sizes to force distribution of the water throughout the block at various speeds. Some engines may not show it from the outside, however. And many of those engines have heating issues today because after so many decades, the internal distribution plates have rotted away. On the '25 Pierce arrow I used to have, I took the side panel off the engine, and the remains of the distribution panel came out as blobs of rust. After several attempts to find good information, and finding information about the piece was simply not available, I designed and made my own. Testing the block temperatures in various areas under differing conditions both before and after, showed that my design made a huge difference in how evenly the block was cooled. Uneven cooling is one of the greatest reasons for blocks and heads cracking.
A close friend of mine did the same thing for a '21 Stutz (on it, the original distribution plate was dissolved aluminum). We had numerous discussions about what was involved in distributing a proper flow of coolant. We also had several discussions about how many antique cars out there are running without proper water distribution because hobbyists and restorers think it won't matter to run without the proper distribution panel when they find the original rotted to near nothing.
I suspect that the uneven heating of the engine was discovered really quick, and Ford decided to make the change as soon as possible, which was basically the first 2500 cars using up the first run of engines.
Wish I had documentation to back up that speculation.
Any pics of a water pump from one of the first 2500? Iíve never seen one. Iíve only seen the aftermarket pumps.
So Wayne based on your speculation of uneven heating of the engine would YOU tell people to not install a waterpump on thier Model T?
From the 1908 Autocar article and the contents for this issue...there was high praise for the thermo-syphon cooling system.
If you haven't read the article I suggest you spend some time to see the comments from 1908 about the system. The article discusses race cars of the period using the system successfully. Was not Ford heavy into racing by 1908-1909?
A point in the article was...operating temperatures are reached quickly.
https://books.google.com/books?id=4n8fAQAAMAAJ&pg=PA284&dq=thermosyphon+system+1 908&hl=en&sa=X&ved=0ahUKEwjfgpnGyp3XAhXp5oMKHW5aAlwQ6AEIJjAA#v=onepage&q=thermos yphon%20system%201908&f=false
Thank you so much for your thoughts. And more area always welcome.
George -- by any chance do you have some additional details on where the "letter was found in the back of a ledger from the Ford Company dated 1908 1 April?" Benson Ford Archives? Other? Can we possibly obtain a copy etc.?
And thank you for the link to the 1908 Autocar article -- it was great. Clearly he supported thermo-syphon and shared valid reasons why.
Gary -- please Google water pump 1909 engine site:mtfca.com and lots of information comes up. Some of which I will try to repost below.
Wayne thank you for your thoughts. Looking at the photos of the water pump engine -- the cooling may have been better than you originally thought. Note that the there is not a water passage at the very front of the engine from the block to the head like in the thermo-syphon engines. From:
http://www.mtfca.com/discus/messages/118802/128695.html?1267915959 Ralph Ricks (RIP) posted :
Dan Treace posted the photo of the head. I rotated the photo 180 so the front of the head would be to the right on the photo. Note there is NOT a water passage at the front. Note also there are only small holes between the cylinders and not the large ones used on the later blocks and heads. The last rear water passage is large like the thermo-syphon one, but it is mostly not visible in the photo.
Again thank you all for putting some thought into this.
Hap l9l5 cut off
Hap, The problem is still that all the cooled water surrounds number one cylinder, then moves down the line past numbers 2, 3, and eventually 4. Then moves up into the head and all the water runs up the line past all the combustion domes. By the time number one is reached (over which the water is sent up into the radiator), the water is so hot it almost cannot cool number 1's combustion dome enough. This creates a difficult problem for the head. The front of the cylinder block is kept cold, only a few degrees higher than what comes out of the radiator after being cooled (usually about 140 degrees) while the front of the head is near boiling (probably about 200 degrees if all is well). The expansion differential between the front of the block and the front of the head due to the temperature difference is tough on iron and gaskets.
To cool an engine properly, the cooled water (coolant?) needs to be divided as it goes into the engine (that being the purpose of the rotted away distribution plates on the Pierce and Stutz mentioned above), then flow slowly up and recombine as it goes into the radiator. There will always be a temperature differential between where the water goes in, and where it goes out. But you want the differential from the front and back to not be extreme. Overall, the blocks temperature at a proper operating temperature should not vary greatly. There should not be hot spots and cold areas. That sort of differential in expansion is what causes head gaskets to fail, and heads or blocks to crack.
I did recall that there was little to no passage on the front of the early engine, but did not recall just how much there was in the middle. Thank you for the pictures.
I recall from reading about the famous camping trip taken in the pre-production (called by some historians as model T "0") by Henry Ford and a couple other engineers, that they had several issues with the car. One of the oil leaks was solved by breaking off an oil line that was intended to oil the "fourth main bearing" (and quickly deleted from production). Water pump leaks were also mentioned, as well as unspecified cooling issues.
Again, pure speculation on my part. But I wouldn't be surprised if they started designing the new improved model T motor right away.
At that point, the first engines were probably already being cast and built. And it would have taken a few months to have made the revisions, pattern making, mold making, and begin casting and machining.
Doug K, I mostly try to stay neutral on the water pumps on model Ts issue. I am not a big fan of them, nor do I hate them. One of my long-time best friends is adamantly opposed to a water pump on any model T (except of course for historic purposes on the first 2500). He has been known to buy a model T, and the first thing he does when he gets the new toy home is take off the water pump and hit it with the cutting torch!
The water pumps are mostly little to no help. In a few cases, they may help just enough to keep a bad radiator going awhile longer. The fact is, many model Ts will run too cold with a water pump. Thermo-syphon is a wonderful self regulating temperature control. Most water pumps made for use on a T engine push the water into the inlet (middle of block) and force the water to flow on up and out without regulation of temperature. I have checked with my little laser-pointer thermometer and found the front and back of model T blocks with water pumps to be too cold. The beauty of the thermo-syphon is that the warmest water rises fastest, allowing the colder areas to warm up to a better operating temperature before that water exits.
Nearly all model Ts? The number one cylinder runs a little too cold (a combination of coolish air blowing by from the radiator and being the shortest least resistant pathway for the coolant to pass through the engine circulating water faster by number one). This problem is worse with a pump.
Many people recommend using a thermostat if you have a water pump. While this will raise the overall temperature to a better operating level, once the thermostat opens, the water will still be pushed through the shortest and least resistance paths leaving different areas of the engine at different temperatures. As long as the mix of too hot and too cold are high enough to keep the thermostat open? The flow will continue, and continue unevenly.