Farmall 400 2-71 Repower Project (Step by Step with Pictures)
Update Oct 14, 2018: I’ve made recent upgrades to the engine and have updated this “Step by Step Story” to reflect changes made since my original posting. Basically I ended up with coolant in the oil and head gasket compression seal leakage into the coolant. Modifications were needed. My latest discovery relates to a new finding about the original head gasket being significantly thinner!
Background: Head gasket upgrades were made to Series 71 engines, in the early years of production. I do not know the exact year of the change, but a service manual for 6-71s indicates the transition serial number for the 6-71 was 6A16283. Considering the number of 6-71’s built in WWII (http://usautoindustryworldwartwo.com/General%20Motors/detroit-diesel.htm ) and that 6A472101 was built in 1995, and 6A 90964 was built in 1960, it would appear the change was made in the early or mid 1940’s. A 1944 Gray Marine Manual I have still only shows the original low block design. All series 71’s were upgraded with the exception of the 2-71, which retained the original 1938 “low block” design until it went out of production in 1987. The new high block design used individual compression gaskets and “o-ring” seals for oil and coolant versus the original 2 piece body gasket. The result of the new design was that much more of the head bolt loading went to the compression gaskets, or at least in a more controlled manner. I assume the changes were made to facilitate the use of larger injectors, power ratings and or turbocharging. For example 90 cubic mm injectors were used in high output 6-71s in WWII. All 2-71’s had 60 or 70 cubic mm through out the production period. It was the only engine in the series (other than the 1-71) not to be eventually turbocharged or fitted with the higher compression ratio 18.7:1 pistons in non turbo configurations.
My failure analysis shows my original J-B weld repair of the erosion around the water holes in the block did not hold up well. It is also likely the higher cylinder pressures from the 18.7 compression ratio pistons and the added boost (2 inches Hg) due to the coated blower, in combination with larger output injectors caused higher cylinder pressure resulting in the compression gasket leakage (bubbles in the radiator). I also later discovered the aftermarket head gasket I used was roughly .015 inch thinner and only had 4 vs 5 steel layers. The effect of this would have been to raise the nominal compression ration to roughly 19.6:1 further increasing cylinder pressures! This was only identified after removing the second gasket and comparing it to the 2nd and 3rd. More on this follows.
My initial fix pursued was machining steel inserts for the block to fix the erosion and changing to 17:1 turbo pistons and rings to decrease cylinder pressure. Pictures showing the modifications to the eroded water holes appear later in the picture story. The liners were shimmed to near .006 inch protrusion to increase the load on the compression seal. While the engine runs good with the 17:1 pistons, I decided to go back to the 18:7:1 compression ratio and perhaps limit the injector size and or retard injection timing, limiting the maximum rpm and maybe increasing head bolt load to deal with cylinder pressures. My reason for changing back is the startability of the engine has been greatly diminished and the start up white smoke is much worse.
Previously the engine had almost instant starts with virtually no white smoke even at around 32 deg F. With 17:1 pistons it white smokes for a significant period of time after starting even at above 70 deg F. This morning on a 42 degree cold start I need to use ether and the start and clean-up was not impressive. With the white smoke (unburned fuel) it is a nuisance when starting inside my garage/shop. I’ll publish more on this in the future.
In studying injector combinations I was able to obtain information on the injector P&B (Plunger and Bushing) assemblies I had experimented with 5228658 (N75) and 5228661 (N80). N-80 has 4% higher effective stroke or theoretical output than the N-75, which I previously understood, however I found the N-80 has approximately 3.6 degrees of additional timing advance build in! At 1.460 injector timing height, the theoretical beginning of injection would have been at approximately 20 degrees before TDC, which is quite advanced. One new possibility is that my brief experimentation with the N-80 may have contributed to higher cylinder pressures and the head gasket compression seal leakage!
Since rebuilding with new 18.7:1 pistons and using the thicker OEM head gasket, the startability is restored and there is clean-up of the white smoke within seconds or starting. There is a small amount of white smoke for a few seconds not previously there with the first build. I contribute this to the first build probably being closer to 19:6 :1 due to the thinner head gasket. I also sense a bit more black smoke when highly loaded. The lower smoke with the thinner gasket might be due to the lower piston to head clearance, which give more squish velocity and a higher percentage of air in the bowl at TDC (higher k-factor).
I saw some suspicious wear or coloration near the top of the liners, initially I thought this may have been due to the coolant in the oil and or in combination with the higher cylinder pressures and or insufficient lubrication. Detroit diesel used different oil ring packages, typically high tension oil rings for injectors below 60 mm3, reduced tension oil rings for non turbo engines above 60 mm3 and a single lower oil scraper in the lower groove for turbo engines with larger injectors. To be most robust relative to lubrication, I used the turbo oil rings in my 2nd and 3rd builds due to the high output injectors I’m running with the increased boost. Upon tear down of the 17:1 build, the wear marks and coloration appeared similar. My conclusion is it may have been normal as the fire-ring had not yet fully seated. For both the 2nd and 3rd builds I did a slight re-hone and the areas cleaned up quickly.
Picture Story of the rebuild:
I primed the fuel system until fuel filters were full and fuel was coming out the return line. The engine fired right up with only a few turns of the crank. The above video is a typical start. I found the used old water pump and fuel transfer pump were both leaking so I have some work to do. A replacement water pump was $782 exchange so have decided to rebuild it myself for about $132 of parts. New fuel pump seals will be about $6.00.
I needed to connect connect the engine throttle lever to the tractor’s driver control lever. To achieve this I constructed a cable system where a 1/8 inch stainless cable runs from the engine’s lever to the lever on the control shaft on the opposite side of the engine through a piece of heavy wall fuel injection tubing The direction on the operator control shaft is reversed,, with up being maximum speed/power.
I plumbed the hydraulic system but had problems with the hydraulic system overheating if I increased the RPM. I added a a pressure gauge and found that at about 800 rpm the pressure regulating system jumped the pressure to around 500 psi and the regulator would not release when RPM was dropped unless I moved a control valve. I concluded this was a problem related to the very large hydraulic pump I had used from the original Allis Chalmers HD5 dozer engine. It took some research to understand the Farmall hydraulic system to decide a course of action.
The Farmall hydraulic system uses closed center hydraulic valves and uses a separate pressure regulating system that is engaged via a 3rd hydraulic control passage to the control valves. When a valve is moved, the low background pressure in the control passage is dropped and the pressure regulator raises the pressure. When the valve is returned to it’s center position the pressure is reestablished and the pressure regulator reduces the working pressure to a low level. With the very high flow rate, of the HD5 pump, the pressure drops across the pump was causing the regulator valve to raise the pressure. I briefly explored the possibility of increasing the size of a control orifice in the regulator block to try to keep the regulating system from malfunctioning, but ultimately decided to find a hydraulic pump closer to the original Farmall pumps size.
The pump I chose is a model GP-F20-25-S9-A from Dynamic Fluid Components Inc. rated at 2900 psi. The Farmall hydraulic system is regulated in the 1200-1300 psi range so the pump provides good factor of safety. I purchased it for $110 from Wholesale Hydraulic Warehouse. The pump has a displacement per revolution of 1.52 cubic inches which I estimate is smaller than the original Farmall pump that appears to be rated at 12 gpm at 1450 rpm (1.9 cubic inch per revolution). This smaller pump would have a flow rate of about 12 gpm at 1800 rpm, where I could run the 2-71 for maximum power, and avoid any chance of the system malfunctioning and overheating.
Installation of the pump required the fabrication of an adapter bracket, the modification of the drive coupling and a resizing to the attachment fittings. The revisions are shown in the following photos.
Latest pictures showing present state of the project follow, sheet metal work to be completed in the spring:
The fast hitch assembly was also in dismal shape and required extensive repair. Besides the normal rust and wear and tear on mounting bracket was broken in two and had to be welded and the one on the opposite side had been previously broken in half and re-welded withe the two halves improperly aligned. Two of the mounting holes had been torched to an oversize to allow assembly without proper support of the bolts. The following slides show how this was repaired by mounting the part, inserting a carbon stick I had threaded with a die into each hole and then welding around the carbon stick to locate the carbon stick. The bracket was removed and the carbon stick reinserted and the balance of the oversized hole was welded shut from both sides, the carbon stick removed, the surfaces ground flat and the hole lightly cleaned up with a die grinder.
In addition to the bracket repair the hydraulic cylinder was rebuilt with new seals and the drawbar, which had been broken in the middle and welded by a previous owner, was reinforced by adding a home built extension plate. This also located the hitch point to the industry standard 14.5 inches beyond the power take-off shaft.
After several months of running I found coolant in the oil and air bubbles in the radiator. Suspecting either a head gasket issue or a cracked head, the head was removed for inspection. Deterioration of the JB Weld Fix of the erosion around several of the water transfer holes had failed, as shown in the above picture. Compression gasket leakage was also observed.
Engine block deck with machined inserts set in place. Each was surface ground for flush mounting. Anaerobic sealant was used under each insert just prior to assembling with a new head gasket. Note, liners and pistons were removed for the modification.