Rebuilding Cylinder Heads

Rebuilding cylinder heads is not particularly difficult and may be done in the home shop with a lathe, mill, a tool post grinder and a few "shop made tools." I have brought several engines back to life for little cost by doing the work here at the house. In this case it is a 1989 Nissan Sentra with a 1600 cc over head cam engine.

On our evening walks around the neighborhood I noticed a car had been sitting in a neighbor's yard for several months, maybe a year or more. One day I happened to see him outside and asked about it. He said that the car had overheated and despite replacing the waterpump, radiator and hoses, it still blows all the water out after only a few minutes running. I suggested that the head may be warped and that car may be fixed. He said that he had already bought a new car and only wanted to get the Sentra out of his yard. I told him I needed a 4 door car, I could probably fix it and asked what he wanted for the car. He replied $50 to $100. I offered him $75. He said, "sold!" I brought a portable air tank over, filled up the flat tires and with a few friends, pushed it home.

Before removing the head, all the connections are labeled. It is also a good idea to photograph all the hoses and wires if you have a camera. All the accessories, cables and the valve cover are then removed. A spot on the timing chain and sprocket are cleaned and marked with a "Sharpie" pen. The cam sprocket is removed and the chain is secured, The "tightening – removing" sequence of the head bolts is found in a Haynes service manual available at auto parts stores. The head bolts are loosened in steps, about a quarter turn at a time. When all the bolts are removed, the head is rapped with a rubber mallet (shot filled dead blow hammer) to loosen it. The head is lifted off the engine and set on the bench.

Before any judgements are made as to the condition of the head it must be thoroughly cleaned. The best cleaner for this job is "Purple Stuff" or the equivalent water –based product sold at auto parts stores. It REALLY WORKS WELL, much better than the solvent based products I have used. When the head is clean, dry it with compressed air and give it a light coat of WD-40 to prevent oxidation. Carefully scrape the remaining gasket with a gasket scraper being careful not to nick or scratch the head. Wipe or rinse the head clean again.

I carefully look for cracks, especially in between the valve seats. The head looks good so far, now I'll check for warp with a straightedge and a feeler gage. The head had .010-inch warp around and between cylinders #2 and #3. Other than that, it looks pretty good. The maximum total height that may be machined off the head and block combined is .008 inch, however the specification allows .002 inch warp. Machining more than this brings the cam closer to the crank, changing the valve timing, among other things. There are a number of options here. I can machine off .008 and live with a little bit of warp or I can completely clean up the head and get a copper shim to restore the original height. Being cheap (it was a $75 car) I chose to live with .002 warp in the head.

I remove the valves with a Craftsman valvespring compressor. Because many valve springs are wound with variable pitch, I make a note of the coil spacing. I put each valve and all the components for that valve in a labeled ziplock bag. The valves are cleaned up later, after I see how well the head machines. I lightly sand blast the carbon from the combustion chambers and the ports being careful not to get sand down in the valve guides where it may become embedded in the bearing surface. It also a good idea to try to blast away as much of the deposits in the water jacket as you can. This head is warped because it over heated. Try to eliminate as many of the possible reasons for the overheating as you can. Clean water passages can't hurt. I blow all the sand out with compressed air and take the head to the mill.

The mill tabletop is cleaned with kerosene. Any burrs, if found are smoothed with a fine wet-stone. The mill table is wiped clean again and then I run my bare hand over it to detect any remaining foreign material. The spindle is "trammed" or set so that it is perfectly perpendicular to the tabletop so that successive surfacing cuts across the cylinder head will not have a burr between them.

Tramming the spindle is done by placing a set of precision ground parallels on the table and checking the height of them with a dial indicator. They should be the same height in all 4 directions (left, right, in front and behind the spindle). The dial indicator is mounted in on a tramming bar that is chucked in the spindle. Although a long bar is best for aligning the spindle, I have had good results with a Starrett "last word" indicator and the short bar that comes with it. I am using a very flat and parallel section of a disc (Photo1). I happened to come upon in the scrap yard one day. It is pretty easy to align the spindle with this set up. When tramming the spindle, don't forget to clamp the knee down. This is sometimes overlooked producing inaccurate results. This can be a time consuming job, but it is well worth all the effort you put into it. Your cuts will be much smoother afterwards.

No matter what kind of head you are surfacing, the real trick is to clamp it to the table without distorting it. This particular head is machined parallel on the top and bottom. I often clamp through the sparkplug holes when surfacing "flatheads" however the sparkplug holes are at an angle in this head, so that is not an option. The cam bearing journals are all machined flat on top and they are level with the top surface of the head. I choose to insert a 1½-inch diameter bar through the cam journals and clamp it as seen in (photo2). Before clamping, I drill two holes in the bar, carefully clean it and remove any burrs. I want to be sure that no dirt is on the bar so that it does not become embedded in the cam bearings. Because the head has a little warp, and because I want a direct clamping load (not transmitted through the head to another surface on the head) I insert parallels under the journals. The clamping nuts and bolts are lubricated and torqued in steps to 20 foot pounds. The trick here is to securely clamp the head, but not to apply enough force to distort the head.

I am using a 2-inch carbide cutter to surface the head. It is a little slow because I have to make several passes to cut the whole width of the head. I start by setting the cutter to height with a sheet of paper (photo3). This paper is .003-inches thick so I raise the table until the paper just catches. From there I remove the paper, raise the table .003-inch and set the dial on zero. Aluminum may be a little gummy to machine so I'll use a cutting fluid. I often use kerosene, but because I want a really smooth finish, I use "tapmagic." It produces an almost mirror finish. (photo4). Good surface finish is important on aluminum heads because, when at operating temperature, aluminum expands at about 1.7 time the rate of cast iron. The expansion and contraction of the head combined with a rough surface saws into the gasket material leading to early failure. Cast iron heads do not need to be as smooth because they expand at the same rate as the block. I make cuts .002 deep, use a spindle speed of 1340 rpm and a slow table feed to cut the surface. My final passes are .001 deep. The head cleans up well, only 2 small places are left un-machined. I can barely get a .001 feeler between a straightedge and the low spots. I am happy with the work so I loosen the clamping bolts in steps and remove the head.

Valve seats are cut with a grinding stone available from autoparts suppliers. I am using a 1 ½-inch diameter general-purpose gray stone made by "Goodson." It is made around a brass center threaded 11/16ths 16tpi so that it screws onto a small motor. Using a lathe, I made an arbor to hold the stone (see drawing). When cutting the seats, I can turn the stone by hand or use a small drill to drive it. The arbor is cut from a 6-inch length of ¾-inch steel rod. The shaft is knurled to make it easier to turn by hand.

The really critical part of the arbor is the insert that centers the tool over the valveseat. The inserts are machined to the same diameter as the valve stems on one end and the other is machined to a good fit in the arbor. When inserting the tip in the tool, airpressure gives some resistance and when removing it makes a popping sound somewhat like a cork. I get this fit by drilling and reaming the arbor to final size (5/16-inch), this give a smooth finish. I then turn the inserts in the lathe and polish them at high speed with sandpaper until they have a smooth fit with minimal clearance. Many heads have different diameter intake and exhaust valve stems. I had to make two inserts for this head (Photo5). SETTING THE LATHE UP TO CUT VALVES AND SEATS: The valves are ground in the lathe by chucking them in a dedicated chuck and cutting the face with a tool post grinder. I have a set of collets that I use to chuck the non-metric valves. I use a chuck setup for all my other nonstandard valves. I am using a ¾-inch capacity ball-bearing chuck available from ENCO. The chuck is mounted on a 5MT taper adapted that fits the lathe spindle. I chose this large combination because I wanted to drill out the center of the taper adapter so that the valves will fit down into the spindle. My taper adapter is drilled out ¾-inch diameter and about 4-inches deep. This is easy to do, just put the adapter in the lathe spindle (without the chuck) and drill it using a drill-bit clamped in the tailstock chuck. The large chuck and adapter were fairly inexpensive. Realistically you can get by with a smaller chuck and drill your taper adapter out to 3/8-inch diameter because most valve stems are 5/16-inch diameter or smaller.

Valves usually are ground to 45-degrees. I have had some valves that are ground to 30-degrees, you will find this on old flathead Continental engines that are used in old Jeeps and fork-lifts. To improve the flow, some valve seats have 3 angles, 30, 45 and 60 degrees. If you want to cut all three angles you will have to have three stones. Set the compound rest to whatever angle that you need to cut. Both the valve and the seat cutting stone should be ground without changing the setting of the compound rest of the lathe, so that the seats and the valves have the identical angle.

Carefully clean the lathe spindle and adapter to be sure there are no particles on either one. Any dirt here will destroy the accuracy of the valve grinding operation. Mount the tool post grinder on the compound rest and true up the stone with a diamond wheel dresser. (Be sure to cover the ways of the lathe with cloth or paper to protect them from the grinder dust.) I made a simple dresser holder from a section of ¾-inch diameter steel rod (photo6). It is drilled at 90 and 45 degrees and also has a set-screw on each side to hold the dresser in place. The dresser may be chucked in the head or tailstock for cutting the wheel. In this situation it happens to be in the headstock (photo7). I have the spindle set in a low gear so that it does not rotate when I cut the wheel. For the best accuracy, the bearings in the tool post grinder have to warm up so I turn it on and let it run for 10 to 15 minutes before making any cuts.

Chuck up an old valve as close to the head as you can as shown in (Photo8). Set the lathe to run about 300 rpm in the reverse direction. Make a practice or "test" cut using the compound rest to feed the grinder across the valve. I have a good supply of old valves and springs I got from the scrap yard. I took my spring compressor down there and stripped several heads. Steel was selling for 5-cents a pound so I may have a dollar or two in all of them.

When you are satisfied with your grinding set up, start working on your valves. All these valves cleaned up with .0015-inch to .003-inch infeed. The whole process may take 15 minutes for all twelve valves.

The valve seats in this head are hardened. They are dirty and only slightly pitted so I'll clean them up by hand (Photo9). Although It may take me 45 minutes to an hour to clean up all 12 seats, I feel like I have more control over light cuts when I do not use the drill to drive the stone. The intake valve seats (Photo10) are cleaning up nicely as may be seen when comparing them to the intake seats in the photo of the sparkplug tap.

The intake and exhaust valve seats are different diameters. In order to clear the combustion chamber walls, I must cut the diameter of the stone down before cutting the exhaust valve seats. I return the tool to the lathe and cut the diameter down .100-inch using the diamond dresser in the tool post as seen in (Photo11).
Clean out the sparkplug hole threads by running a M14 — 1.25 tap through them (Photo12). Cleaning out all the trash helps keep the new plugs from cross threading and tearing up the head. Because aluminum (head) and steel (sparkplugs) are dissimilar metals, I always use "Never Seize" on the sparkplug threads to minimize any corrosion problems. The plugs come out easily when it is time to change them.

When all the machining is done, I scrape any remaining gasket material from around the ports and thoroughly clean the head again. The valvestem seals are given a coat of oil and pushed into place. The valve stems are also given a coat of oil and pushed into the guides. The springs are compressed and the valve keepers are given a coat of grease and put into position. The grease keeps them in place until the springs are released. Be sure to pay attention to the springs. Some springs are wound with a variable pitch. The coils that are wound closer together close up and come together when compressed in use. This raises the natural frequency of the spring to prevent surging and resonance. The springs on the Nissan engine are wound with the tight coils at the bottom of the spring.
I clean up the camshaft with the "Purple Stuff" cleaner and blow it dry with compressed air. I chuck the cam in the lathe with live center in the tailstock. I spin the cam at about 300 rpm and use a fine (approximately 400 grit) sandpaper strip coated with oil to polish it up (Photo13). After I have gone over all the lobes, I reverse the direction of the lathe rotation and polish them again. This cleans both sides of the lobes. I thoroughly rinse the cam and give it a light coat of oil to prevent rust. The journals in the head are oiled up and the clean cam is inserted.

The head is ready to install. Now a few notes about bolts and tightening the newly machined head.

The head bolts, the head and the engine block are in effect stiff springs. They deflect when loaded and relax when the load is removed. They store potential energy and create a clamping force when they maintain a certain amount of energy. Typically about 90% of the energy put into tightening a bolt is converted to heat. About 50% is lost due to the friction at the head of the bolt and 40% is lost due to the friction in the threads. This leaves only 10% of the input work as preload or bolt tension. If there is a 10% increase friction because of dry threads or trash in the threads, then the input work is reduced an extra 4%. This leaves only 6% of the total input torque as preload in the bolted joint. A 10% increase in friction results in a 40% loss in bolt preload. It is very important to clean the threads on the bolts (or use new ones) and to clean the threads in the block by running a tap through them. (Some engines, usually very high compression engines, have rounded tips on the threads to reduce the tendency to crack at the threads. These threads require a special tap.) After the threads are clean, you should use a thread lubricant such as Never – Seize or a Fel-pro product.

Make a final inspection of the cooling system before starting up your new head. I traced the original failure in this engine back to a weak radiator cap. The cap let the coolant boil out and when it was low enough the head ran hot and warped. A $3.65 part caused the original failure.

This engine starts right up and runs very well. I thought it would run, but I am surprised how well it runs. With a little troubleshooting, the airconditioner is brought back to life and blows ice cold (very important in Florida). It is time for a paint job and new tires. I make a 320-mile round trip drive to Orlando several times per week and the car is running great! I was very lucky to find such a car for $75 and I am very happy with the end result. If you want to try cutting a head or two, go get one from the scrap yard and practice on it before you tackle your wife's car. You can probably get an aluminum one for 50-cents a pound. Once you have done it, some of those other restoration or repair jobs are not so intimidating. Good luck!

If you are interested in more information on the statistics of bolted joints, then see "An Introduction to the Design and Behavior of Bolted Joints" by John Bickford. Try a school library or request that your local library purchase one.

Affordable machine tools may be found at Enco. Phone number 1 800 USE ENCO. Web address:

Your IP Address is: