Valves, valve seat cutting and valves

[gbritnell]

Gentlemen,

I didn't want Brian's thread on the Kerzel build to get too far off track. I know there is a discussion about valves and valve sealing but rather than add to it I thought I would post a little of my findings on a new thread.

First to answer Nick's question about just knocking the edge off of a square seat, the answer is yes, but let me present some of my information.

As a lot of members are finding out it's quite tricky (in some cases) to get the valves on an IC engine to seal properly. There are generally 3 methods of machining ports and seats. The first method would be to machine a hole for an inserted valve guide and then the port. The second would be to machine the valve guide into the head followed by the port and lastly to machine a pocket in the head to accept a valve cage.

The problem with all of them is that the initial hole is generally done with a drill followed by a reamer for the valve guide or valve guide insert. Herein lies the problem. There is no way to guarantee that the hole will be concentric to the desired center point. Now depending whether you use a new drill or a resharpened drill for the pilot the hole could be very close or a little off. The reamer will only follow the drilled hole.

As a remedy to this problem you say well then I'll make valve cages separately and press them into the head. You still have to do the same operations on them as you did when you machined them into the head.

Ideally a cutter should be made that would incorporate a reamer for the valve guide at the tip and farther back a counterbore that would then cut the port opening. You would naturally rough out the holes and then follow up with this tool. It would have the rigidity to keep everything in line (hopefully).

I don't do valve cages, why, because I think it's just to much more work. If I have the tools to do the job why not just do it all in one operation?
Now for the fellows that do it this way don't get your hackles up. If it works for you by all means continue doing it.

Here's what I do. Once my initial guide and port are machined in I make up a seat cutter from drill rod that incorporates a stem that fits very closely into the valve guide hole. Above that I enlarge the tool to just under the diameter of the port. This way you won't get any flex if you are using very small valve stems. Now above the larger port diameter I machine a 45* angle out about .08 larger in diameter. This is then fluted and hardened. If you go to page 3 of this thread you will see a picture of the cutter I'm describing.
http://www.homemodelenginemachinist.com/index.php?topic=6205.30
To cut my seat I put the cutter into a small drill chuck, oil the stem, insert it into the guide/port and then just with the weight of the chuck rotate it by hand until I get a nice 'small' seat. I have found that you only need to break the edges on the port to get a good seal.

Now the next problem area, the valves. First off I have made different styles of valves and from several materials. You can use drill rod or stainless steel or for that matter CRS. I have made them from one piece and I have hard brazed caps on the stems. Whichever way you choose to go the critical step is making sure that the stems and heads are concentric, not close but real close.

To machine a valve from one piece, even ones with very small stems I start out by just turning enough material to cut a retainer slot into it. (see Steve's V-8 build). Some valves I don't use a retainer but rather thread the end of the stem for a screw on retainer. I keep cutting back toward the valve head in small steps so that I maintain the rigidity of the stock. Now at this point you could make up a small cup to fit over the end of the valve and put your live center into it to help support the stem but I have never found this necessary. I just continue to cut to my dimension each time and then finally polish up the stem with fine emery paper. I stay away from the head whatever amount my fillet is going to be. I then go in with a tool that has the desired radius on it and cut the fillet. I have my compound swung over to 45* and cut my seat area. With this compete I cut the valve off leaving it a little long so that I can face it down to the proper thickness later.

To make 2 piece valves, the ones with the brazed on heads I make up my stems and heads, braze them together then mount them in a collet. I then indicate them until the indicator doesn't move, at all! Now if you don't have collets here's another way to make a holding device.

Take a piece of brass rod the diameter of which can be about .188 or so larger than your valve stem. Mount it in your 3 jaw chuck and mark it's location relative to the #1 jaw. The reason for this is because even though a chuck won't necessarily repeat for different diameters it most always will repeat for the same diameter. Now carefully drill and ream a hole the size of your valve stem. I then take a very small slitting saw and cut into one side of it to split it open. You now have a collet. I have a small container of these 'collets' and even though my Craftaman 3 jaw chuck has been used for years when I put them into the chuck they are usually within .0005

Now chuck your valve up as close to the collet as feasible to get a tool between the head and the collet and machine your fillet, seat and face. 'TAKE SMALL CUTS'

As a final operation, I slide the valve into the guide and grip it with a small drill chuck from the spring side allowing enough length to let the valve move up and down a little. I charge it with a little bit of very fine valve grinding compound and give it a few turns, back and forth and up and down. Then take the valve out and with a magnifier look at both the seat in the head and the surface of the valve. If they don't look uniform repeat the process until they do. If everything was machined well it should take very little lapping to get them to seat evenly.

I hope this helps some of you who are having issues with this part of IC building.
It sounds like a lot of work but I find that these operations are the most critical to getting a good cylinder seal.

George

 

[bearcar1]

Great play-by-play there George, thank you so much for sharing your methods. Sooner or later (probably later) when I get back on my engine, I'll need to have some semblance of an idea as to how to tackle what appears to be a simple yet daunting task of getting valves to seal properly. This explanation helps tremendously. David Kerzel himself even says to do all of the steps you describe in one set-up, although he does not outline the process in detail like you have and I would have to agree with you about working directly in the head as opposed to implementing a valve cage arrangement as well. The same steps still have to be followed anyway so why go through the gyrations by adding another 'finicky' piece to the mix. Thanks again your write ups are always well received and appreciated. :bow:

BC1
Jim

 

[ghart3]

Good methods and write up and thanks for making it in a new thread so it easier to find for a reference.

 

[cfellows]

Great write-up, George. One thin occurs to me about valve cages. That's what I ended up making for my 2 cylinder open column engine, but a later found that the valve seats were not perpendicular to valve guide. In retrospect, I think my error was in drilling the original 1/8" hole all the way through, which was about 1.25" long. The 1/8" bit must have wandered off center before getting all the way through the blank.

I now think I would have been better off drilling in only to the depth of the port with the 1/8" bit, then opening up the valve port to the finished diameter by step drilling, and, finally, going back in and drilling and reaming the valve guide hole. I think there would have been less chance of the bit wandering off center.

I've made lots of poppet valves in the past and this is the first time I've had this much trouble getting one to seat.

Chuck

 

[Brian Rupnow]

George---Thank you very much. A very informative post, and I hope it will help me out. I may have a go at building a tool like you show in the link you gave. Tomorrow I may just make up a blank plate drilled to match the cylinder head bolt pattern and put it in place with a gasket to confirm that I have compression with the piston and rings alone, irrispective of leaking valves.---Brian

 

[Rustkolector]

George,
Excellent write up. I haven't built many IC engines, but I have learned that my good looking valve seats can still leak enough to cause running problems. To test my valves, I started out using a technique that a local speed shop uses to test valve sealing. They assemble the valve and springs in the head, and spray penetrating oil into the port and let the head set for an hour or two. If a seat leaks oil, they redo it. I have tried this method and even with no oil leaks overnight, I still had a poor sealing valve or two when running. I now use a hand vacuum pump like the ones used for bleeding auto brakes. I assemble the valve and spring using #30 oil on the stem and guide. I operate the valve by hand a few times, then pump up to 25" Hg vacuum, and watch the pressure bleed down. If it holds more than 15" Hg for 20-30 seconds, the seat is doing pretty well. An excellent seat will hold 20" Hg for 30 seconds. With vacuum testing, you know precisely how well your valves are sealing. No matter how well you did on the rest of the engine, if the valves don't seal adequately, you will have peformance issues.

Jeff

 

[jpeter]

I test my valves by screwing the head to a plate, gasket and all. I've got a sparkplug, for an overhead valve engine of course, which I've drilled to which I've added a small tube, I use model airplane fuel tube. I blow on the tube while I hold the head under water. My limit is a bubble every 5 or so seconds. Any faster rate of bubbles and I go back to work on the valve. it works for me. Regarding the construction, I have often pressed in valve guides and using a cutter like mentioned before cut the valve seat right in the aluminum. Seems to me the narrower the seat the better. One change I'd suggest is I guide the depth of the seat cutter with my power drill cuz what I've found is often my seat cutter will hog in in spots when controlled only by hand pressure, often to little to see, leaving a deep spot in the seat. A light touch with the cutter is all thats needed.

 

[Drei]

Hi,
I have read a few of youre writings on the valve seat and testing... I chek the seats by pouring some kearsoline in the vavle port. If they leak I do some more lapping with the fine or rough coarse depanding on the ammount that have leaked. I dont know if this method is good enough but kearsoline is a very thin substance thus if it doesnt leak then air wont leak nighter. When I bulit my fist engine and laped the valves this way I achieved a good compression. I would like to know if youre methods are better for valve sealing.

Thanks

 

[jpeter]

Hey Drei, here's my response if your asking me about my valve sealing methods: When I first started building engines I had lots of valve troubles. I had lots of all kinds of fit problems cuz I didn't know how good fits had to be. I was teaching metal shop to high school students. As expected my students knew less than I did and I didn't know much. A student built the first engine and it didn't run at all. I followed suit building one to work out enough issues to get one to run, which after a lot of research and redoing, I did. In doing so I developed some standards for students to follow, one being the valve test. They needed a standard to shoot for so they'd have some success after a year of hard shop work. We made cylinders too by boring and lapping. I've established quality standards for them too. In the 5 years or so we built engines we got about 12 to run. The others generally didn't get finished to the point of running but the parts they made were within my specs. BTW, in 40 years of teaching, building engines gave the students the most satisfaction.

 

[Brian Rupnow]

I've came back to revisit this post and thank everyone for their help. It was only after building a valve seat cutting tool like GBritnells , making new valve cages, using the tool and lapping the valves with first 350 grit honing compound, then 600 grit, then finally toothpaste, that I was finally able to get my valves to seal properly and my engine to run on its own. The valve grinding compound sold by Permatex #474G is extremely course, and for all intents and purposes useless on these small valves. I am in a continual process of learning as I build these small engines, and I cannot stress how vital it is to have a 100% seal on the valves.----Brian