cylinder head

this section covers the head, head bolts/studs, intake trough, valves, rockers, exhaust ports and mods, and pistons. on my 2010 build i did a heavy cleanup of the ports and combustion chambers with a Foredom tool and a big bag of abrasive rolls. the best thing that can be said about this head is that the trough design keeps the ports short and relatively straight. however there was (is) substantial protrusion and sharp corners, and heavy valve shrouding. i was able to clean up a lot of it, but the cylinder head dissection revealed a lot of thin areas that precluded a better job. nonetheless i think it's much improved.

induction (trough plate)

the head has a trough intake, adequate short paths, with only one 90 degree turn each from carb to valve. combustion chamber is a popup wedge. the trough has clever Nash anti-reversion wedges that make for excellent fuel distribution.

some of the intake ports are paired/siamesed, some are not. front to rear, the order is I-II-II-I. this confounds port injection fuel-injector layout. throttle body injection would be adequate anyway.

it has an interesting advantage in that it delivers perfect mixture distribution to all cylinders, a problem on long inline sixes. if you click on the picture to the right, between cylinders 2 and 3 (and 4 and 5), adjacent to the second head stud from the front of the engine, you will see within the right hand trough wall a ramp-shaped protrusion cast into the trough. It pinches mixture flow at that point -- it is an anti-reversion device, preventing back-flow of intake mixture pulses. All six plugs burn to the exact same color.

the trough is covered with a cast aluminum trough plate, a very handy design for hacking induction. It's flat, easy to fabricate from scratch if necessary.

induction improvements

AMC had two carburetor options for this engine: a single-venturi as the base model (Holley 1904, 1906 or 1908 in Americans, or Carter AS or RBS in Classics, depending on year and transmission), and a two-venturi Carter WCD called the "Power Pak" option (relatively uncommon in it's time and now a desirable option). all are very reliable, easy to maintain and un-fussy to adjust and drive, but are lousy performers. for the sort of driving I increasingly do, the WCD annoyingly starved out in every hard turn.

anecdotally, the one-barrel carb "runs out of steam" about 3000 rpm wide open throttle; the WCD will get you about 3500 rpm before it flats out. They are allegedly 200 and 300 cfm flow, respectively, which ought to be enough for more power than that, but experience says otherwise.

initially i searched for carb options to bolt onto the two-barrel trough plate; but I soon realized that the common one-barrel plate with it's standard "YF" bore was far easier to adapt. the Carter WCD shares base pattern with the Rochester 2G (the "small rochester")

The Weber 32/36 DGEV

Redline makes a "Jeep Weber kit" that simply bolts onto the 195.6 OHV's trough plate. the necessary adapter also raises the carb enough to clear the valve cover (which can be removed with the carb installed); if you look closely at the picture to the right, you can see how close the carburetor is to the valve cover.

the 32/36 also runs out of flow in the 3000+ rpm range. however it's tiny primary bore is a great match for low speed luggy stock engine and gears, and is far more tunable than stock carbs. it also does not starve out in turns.

here's the overall relationship of head with the trough plate, adapter and carburetor installed:

Weber 38/38 DGV

the 2017 build required more carburetor, and research showed that the Weber 38/38 ought to be adequate. never before have i had a single component change make so much improvement in overall performance. however, with this build i had also changed transmission (from two different 3-speeds to a custom T5 and carefully selected axle and tire size) and so RPM range and driving habits changed to match, which also allowed for much more aggressive spark advance. all told the old skinny carburetor was clearly holding it back.

the 38/38 however is a drop-in, bolt-on compatible with the 32/36 and an immediate improvement. of course with both carbs you have to fabricate throttle linkage.

(in my opinion, one of these carburetors plus an ignition upgrade would be a noticable improvement at least cost and increased reliability.)

blending the adapter and trough plate

it was easy to blend the adapter to the trough plate and remove all the right angles and sharp edges in the flow. The first thing was to add locating pins to the plate and adapter so that it would stay in alignment after blending . (the adapter has slotted holes that let it slide around.) unfortunately I neglected to take any photos of this, but it's not rocket science. I picked a likely place for two 1/8" pins on the plate where the carb goes and drilled two 1/16" holes about 3/8" deep. into these holes I dropped nipped-off wire brads with the points sticking up above flush approximately 1/32". I carefully aligned the adapter by hand, set it onto the plate (sitting on the brad tips) and whacked it hard with my hand, pricking the bottom of the adapter. I used the prick marks as guides to drill 1/8" holes for the pins, and replaced the brad tips with 1/8" wire. the adapter was to the plate and filed to match.

with the adapter joined to the plate, I smoothed them where i believed the flow would actually go. the big hole in the bottom of the plate had sharp 90-degree edges; I smoothed these as much as I dared, probably a 1" radius. the carb end of the adapter I made slightly (1/32" or so) larger than the carb bore. the inside of the adapter itself was cast lumpy; I smoothed that out substantially, left it with a faintly venturi shape, and increased the overall diameter about 3/16".

avoid resurfaced tappets

for my 2010 build i bought used but resurfaced tappets. this turned out to be a mistake, as the resurfacing apparently removed all of the hardened surface of the mushroom head, precisely where it is most critical. the working face of all 12 tappets was severely pitted, while the matching cam lobe was essentially fine. this was not an oiling issue, but metallurgy.

in the first image below you can see one of the cam lobes; they all looked like that. that is in fact the cam that ran with these tappets.

for the 2017 build i bought new, NOS, tappets from Kanters. they have a faint crown. the mushroom head is black oxide, as is the pushrod ball socket. overall length is 1.878", i measured 6 of the new ones, all were short. i kept notes on two of them, which measured 1.864" in height, or .014" removed, assuming all tappets are initially the same height.

there's a discernable (thumbnail) wear pattern on the cylindrical section, up under the mushroom. this appears to me to be the lateral force caused by the cam thrusting the follower sideways during normal operation; that's a feature of the mushroom design (and why AMC and everyone else ditched them). as the cam operates, the lobe wipes the tappet crown, slightly off-center, rotates the tappet presumably to distribute wear. the follower "rocks" in it's bore with each cam cycle a tiny amount. this scuffs the cylindrical portion of the tappet top and bottom, the rocking pivot point more or less in the center, but offset towards the pushrod end. it's just the sum of the leverages. this leaves (microinches) of gap, where there's no contact, the metal is darker (in the pic above). none of this is unusual or bad or anything, it's just what happens to metal in use. but after 60 years it's a lot of wear.

i stringently ran Mobil 1 15W-50 after a generous dino oil breaking, with ZDDP breakin additive. two changes. that worked great on the 232 (350,000 miles). maybe that's not so good here.


pete had custom forged pistons made that accept a modern 81mm ring set. the dull aluminum piston in some of the photos is a 1970's Silv-O-Lite .060" over piston from a set i purchased on ePay years ago and never used.


you'd think the stock valve setup would at least be a no-brainer here, but it wasn't. the previous machine shop botched this, including resurfacing new valves i had bought in 2010, and cut the exhaust seats too deep. pete had to use a larger exhaust valve, which didn't exist, so a Chevy valve was turned down.

you might also think that stock-replacement valve springs and retainers would be easy, but no: there are different springs for the OHV and flathead versions of this engine, reasonably enough; the flathead spring is 40lbs closed, the OHV is 80lbs. they look identical. Kanters shipped me 40 lb springs back in 2010. (my receipt shows i ordered the correct part). Pete measures every spring he installs and caught the error. they were also binding with what appeared to be the stock retainers. new springs and retainers were found (i dont know what the source is) to solve this. after 50 years nothing can be "assumed".


it should be no surprise that there are no (and never were) aftermarket performance cams for this engine. a used cam was reground, but because the base circle is only .020" or so larger than the rough casting increasing lift or duration is not possible. but Pete's cam genius crammed 110 degrees lobe separation into it.

[for future reference: the L-head engine's cam appears to be identical to the OHV cam except for it's profile, and especially, lift. because it does not have the advantage of rocker multiplication, the lobes are taller, and therefore have more metal that could be ground to make a good OHV cam. this needs to be verified.]

rocker shaft assembly

the rocker shaft assembly is straightforward and reliable. rocker ratio is 1.5:1. the shafts wear and new ones are not available. i've had no trouble with the rockers nor the adjusters. new adjusters are available.

head oiling is accomplished by the rocker shaft itself being pressurized, via the front-most stand, up through a port in the head, which is fed by an external steel line (3/16" inverted flare tube down to either the main gallery, or via the front camshaft journal in 1964 and 1965; the front journal on those camshafts are flatted to provide intermittent top-end oil).

alas, the rocker shaft can't simply be inverted to wear the other side, as oil ports are milled into it to lubricate the bottom (loaded side) of the rocker.

for what they're worth, here are some rocker assembly movies taken while i was adjusting the valves.

exhaust ports and manifold

the exhaust side of the head is overall not too terrible, with the sole exception of the carb-heat provision in the center siamesed ports. in 2010 i equalized exhaust ports to make them all equal.

head sealing

this engine is notorious for head gasket failure and subsequent external coolant leaks, water in the oil, all preceded by chronic overheating issues. see the cooling section for a very simple fix to this deadly problem.

before i had worked out the cooling issue to my satisfaction i also replaced all of the head bolts with studs from ARP. after my experience with them here i will probably replace bolts with studs in all future engines.

Studs are superior to bolts for this application. When a head bolt is torqued, it remains twisted along it's length, due to friction in the threads and under the bolt head. Any transverse motion in the head (caused by thermal cycling...) backs out the bolts. With studs, all of this friction is at the top of the stud, which remains un-twisted. Quality and tolerances are better too.

Since ARP doesn't make a "kit" for this motor and my application isn't particularly stressful on the studs, I simply picked their stock parts from the catalog. There are three different stud lengths. They are coarse threaded at the block end and fine threaded at the top. Twelve-point nuts, machined washers and ARP lube was used. Part numbers are below.

Item ARP part number Quantity Location
Stud, 7/16" x 5.75" AP5.750-1LB 6 Through trough plate
Stud, 7/16" x 5.5" AP5.500-1LB 4 Head ends
Stud, 7/16" x 4.5" AP4.500-1LB 5 Under valve cover
7/16"-20 Nut APN12-1 15  
7/16" ID non-chamfer washer APW1316N 15  
Assembly lube n/a 1 Thread lubricant


apparently at the factory the bodies were set over the engine and transmission assembly on the line. preventing easy insertion from above is the front welded-in cross-brace, just behind the radiator top tank. mine had long ago been hacksawed out, as is common. with it out of the way top-insertion is relatively easy. i've added a bolt-in internal triangular brace between the inner fenders and firewall to replace it.

i've installed engines without the head attached (2010) and with head and complete transmission (2017). the latter definitely requires that the hoist have a load-shifting trolly.

ARP recommends three torque/release cycles on new studs. for the 2010 assembly, and before operation, I did four, without the headgasket, since that gets a one-time crush. I left the third torque to set overnight. After final assembly with gasket in the car I measured stretch on one stud at .012" when torque increased from 20 ft/lbs to the rated 75 ft/lbs. Thanks to David Forbes for the measurement suggestion.

The studs were bottomed in the block and snugged up with an allen key two-finger tight, assembled with ARP hardware and lube and torqued in three stages to the rated 75 ft/lbs. I did not start it until the next day.

Upon every retorque each nut rotated the exact same amount. This was good, because two end nuts will not accept a socket when the rocker shaft is installed; I used a box-end wrench and extender and turned them the same rotational angle as the rest did with the torque wrench. (i'v never found a 12-point box end crows foot socket.)

these same studs were used in 2017, and the builder used his own method of assembly.


the factory technical service manual for these engines has the peculiar requirement of cylinder head retorque schedule of check every 4000 miles, and re-torque every 8000, done while the engine is hot. this is just plain weird, and i am convinced this was due to the head expansion/thermal cycling problem designed into the head, and that is utterly negated by the cooling system fix and ARP studs.

in 2018 i am still running the ARP studs installed in 2010, and other than initial break-in my annual torque-checking has revealed zero need for retorquing. without the thermal-cycling problem there is no need to retorque.

my annual check now consists of setting the torque wrench at 60 ft/lbs and simply checking for loose head nuts. none have loosened. since i have to remove the valve cover annually for valve adjustment (nearly unnecessary) this takes little effort.

The third retorque, at 1000 miles, zero rotation. It appears that stud stretch and headgasket crush is complete. [in 2010 i wrote:] I will continue to check it at intervals, but hopefully the need for constant retorquing is over [in 2018 this seems true].

Here are some pics of the studs installed in the lab. this is the 2010 build.

The stock head bolts penetrate the block exactly one inch. Placement isn't that great, at least to my novice eye; some are along casting side walls, and some are in the middle of horizontal spans. Headbolt spacing is wildly uneven, but there's nothing to be done about that. Here are pics of the stock bolts and their protrusion through a section of a junk head: