modified bottom tap

[petertha]

On the head of my radial engine, the intake & exhaust tube flared lips are retained against the internal ledge of a counterbore by a threaded nut. The threads are 7/16-24 TPI. The bottoming tap I'm using has the typical couple of threads chamfered off so my full threads end about 0.1" from the end of the tap, maybe a bit more. I would really like to gain back as much of this un-tapped distance as I can.

I already modified the heads a bit larger in diameter so I can get more thread engagement. I might be able to deepen the counterbore face a smidge more, but then it starts getting close to other holes & features. The sketch doesnt show a .050" washer but I've already got that factored in.

Is it viable to modify another bottoming tap, grind the chamfer off nearly completely & chase this thread further down to the bottom? Somewhere I thought I read this doesn't work very well, but I cant recall if true or why. Very weak tap edge that will likely break?

Alternatively I could turn down the male threads on the nut so it penetrates further in. I just have very few threads to engage the head as it is. Plus the nut enters the head at an angle so its partial hole contact & thread engagement is further reduced by cooling fin grooves.

 

[Hopper]

I can't see why you couldnt grind the end of the tap down as far as the first tooth, after tapping the hole with the tap as it exists now. I've ground the ends off many a tap over the years and never had a problem.

 

[XD351]

I have done this many times also as per what hopper says with no issues, the only other way is to machine a relief in the female thread with a boring head or by setting it up in a lathe and using a boring tool to cut the relief at the bottom of the thread.

 

[mayhugh1]

Petertha,
I also have done this many times. You just have to be careful when starting the re-ground tap in the already threaded hole. The edges of the re-ground tap will be sharp and you'll have no lead-in threads to help get you started. - Terry

 

[petertha]

Thanks for the comments. There is another factor I kind of lost track of. The counterbore is better suited to an end mill over a numbered or lettered drill because the hole is shallow and its entering the head tangentially. A drill would just drift as the cone enters the hole with one only side in contact. Just so happens a 10mm EM is a nice % thread fit to a 24TPI.

If I go to next finer 28 TPI, I get a few advantages. The already thin annular wall thickness of the nut increases a bit, so a teeny bit less likely to crack the nut - thinking the hot exhaust side. I get a little bit more thread engagement along the length although its almost insignificant on this short distance. The pilot hole for 7/16-28 would have to be 13/32. Not sure I have that end mill but I see they are available. Also, now I will go back & measure -I bet my 10mm EM is cutting a bit oversize & therefore might be a good fit for 28 TPI in reality.

All things considered on this particular application, maybe should I opt for a finer thread? can you think of any downside?

Not surprisingly.I'm right in between an M10 x 1 which is what the original (metric) design called for. But I was hoping to stay with imperial threads on my lathe without the change gear rigamarole & generally more expensive taps to sacrifice.

 

[WOB]

I suspect your tube flange design is problematic. When the nut face flare puts pressure on the tube flange it is going to try to swage the flange smaller because of the reverse flare on the nut face. There is nothing inside the tube flare to resist the effect. When the nut flare edge bottoms on the head counterbore, the tube flare is not captured mechanically. You are depending on the inherent elasticity of the alum. tube and its friction against the counter bore bottom to maintain the seal. Soft aluminum creeps under load, especially under vibration and rapid temp changes. I think you would be better off by far if your tube flare was 90 deg and full counterbore dia. and the nut face was also 90 deg. Then the tube flare would be captured mechanically with no place to go and would also provide a solid gas pressure seal. I have built 13 engine cylinders with the aforementioned tube connection with no hint of failure. I used modified SAE 1/4" steel nuts which have a 7/16-24 thread. They look very much like your custom made nuts after trimming the excess metal . Just my 2 cents. YMMV. Good luck.

WOB

 

[john_reese]

Like WOB said.
Look at an inverted flare fitting. All have a projection to engage the inside of the tube flare. Without it your flare on the tubing will crush.

 

[petertha]

Thanks for your input WOB. I do have an example of a exactly what you are talking about (pics). It came off a .60CI size glow 4S and has a right angle lip. It is very thin walled tubing.

I did actually consider bending a right angle flange although I haven't yet tried making that kind of bend in my little fixture. Here is where I kind of got stuck. Maybe check my math if I did something wrong.
The pilot hole for 7/16"-24 TPI thread is say letter X drill = 0.397" gives 75% thread engagement. So that bore diameter becomes the crest of the internal threads & by default the maximum diameter of the upturned tubing thread. If it was a hard-90 deg or a ring soldered on the tubing all would be good. But if I apply any kind of bend radius to avoid cracking there is very little remaining perpendicular flange, its mostly all bend radius. Pic shows TLAR radius same as wall thickness of .028" I've also seen those cool double bend brake line treatments. I dont have my notes in front of me but I thought they were larger OD yet. And the jigs require a longer straight length of tubing in order to apply that treatment.

Another issue maybe not so obvious is the ID of my nut nut is a bit larger because I want to slip in on after the tubing end treatment.If I make it just oversize the nominal tubing OD, I cant get it on or the bend radius has to be quite large.

So that's what got me to angled flare with a matching angle nut. But the face of the tubing is still finished parallel to the face of the counter bore. I was hoping that would mate face to face.

There will also be a washer in there (not shown) to seal but I don't think that is really factoring into what you are suggesting which I think is about a distortion problem by tightening. Now you have me curious, I have another 4S model engine header I will look at now.

 

[petertha]

Just to make sure we are seeing the same thing magnified a bit. Green arrow indicates the face of tubing made flat so it mates flat is & coincident with counterbore face.

 

[WOB]

I see your logic. I got around the problem by placing the nut on the tube after flaring and before bending. I needed a 1/2" radius ( inside of the tube dia.) bend that started immediately as the tube exited the nut. So, my bending jig incorporated a replica of the head recess to hold the flared end of the tube by its nut connection. This means the nut is captured on the flared end after bending, but there is no problem. Your nut ID can be only a couple of thou. larger than the tube OD. You need a small radius on the inside corner of the mouth of the nut. I did it with a deburring tool. 3003 alum tubing is very malleable and I had no problem with cracking when cold forming the 90 deg flare in the lathe. You do not need a sealing washer. Think about it. When the nut face comes down on the backside of the flare, the tube flange is trapped and swaged into perfect, gas tight fit against the head counterbore on one side and the nut face on the other. What else do you need? I am not saying your design will not work, but it seems to me to be likely to loosen up given some running time. Of course, given the fact that our engines rarely see much running time before being retired to the display case, you might be OK. Good luck

WOB

 

[Jennifer Edwards]

I realise this is a bit of a late response but wanted to add my validation,
here goes:

When I purchase “bottoming” taps I always buy two and grind the end of the second one down quite a bit so I can follow up with it and get those few extra threads cut in.

Remember to not overheat the tap when grinding. Use plenty of water/coolant so you do not take the hardness out of the tap.

 

[petertha]

I see your logic. I got around the problem by placing the nut on the tube after flaring and before bending. I needed a 1/2" radius ( inside of the tube dia.) bend that started immediately as the tube exited the nut. So, my bending jig incorporated a replica of the head recess to hold the flared end of the tube by its nut connection. This means the nut is captured on the flared end after bending, but there is no problem.

Thanks again. I would like to see any details of your bending apparatus. You got a 1/2" bend on 5/16OD tube? I'm a bit of a newb in this regard. I bought a Ridgid 36092 which is dedicated to 5/16 OD tube, minimum bend radius is 15/16". Actually they don't really say of that dimension relates to the center line or inner bend diameter or...? I suspect its the center line because my test nuts pretty much confirm.

I am OK with this 'fixed' bend radius on my application, meaning aesthetically it looks nice enough & the tool is quite straightforward to use. But one of the issues is if I do the end treatment first, like a flare or whatever, I have to place that outside of the grasping clamp (green line). That adds about another ~ 0.8-0.9"? of straight segment before the bend radius begins (red line). That actually does not look as nice & creates other issues vs. the flare right on the radius quadrant. So that's where I am cutting the tube & then apply to custom flaring jig. I have seen dedicated souls make custom benders & filled with the soft metal or sand to achieve even tighter bends. I was hoping to avoid that if I could for now.

I worked out the min ID hole of the nut to achieve clearance around the bend & machined some simple collars to confirm this. The longer the nut length, the larger the hole, which means less strength because its closer to the valleys of the thread. So I've tried to keep the length at minimum for this reason. I have made a bigger chamfer on the hex nut side which is another way of saying effectively reducing the length where the tube contacts.

I've got some pics coming for a larger commercial 4S glow. On the exhaust side they threaded the tubing itself & then a simple thin gland nut to retain. I could do that too because the exhaust pipes are just little stubs, but it wont work for the induction side because those are longer pipes that spider their way back to the rear of the engine into manifold - a multi 3D bend type deal.

 

[petertha]

Thanks for the tip Jennifer.
I did some more messing around & I think 28 tpi offers more advantages over 24. So I'm going to order those taps plus a spare (or two!) to grind.
I found an exhaust pipe from a commercial (OS-120) engine & looks to be ~ 11mm OD (0.433") and 0.8mm pitch (.031"). My 24 TPI is .041 pitch, 28 TP is .035" so in the ballpark.
28 TPI will give me more threads engaged over the short (and non-symmetrical) head engagement plus a bit more net wall thickness of the nut which I am concerned of cracking.

 

[WOB]

This is my bending jig.

The 3003 alum. tube has been flared 90 deg. and is held in the "L" shaped block in a duplicate of the cylinder head's recess with a brass flare nut that has been machined flat on it's face. The tube is filled with Cerrobend. The tube is bent to the right 90 deg. over the 1/2 round block ( 1/2" radius). Due to the tube's malleability, the bend is pretty easy and there is no change in the cross section shape. You might need a cheater pipe over the tube to give more leverage since the Cerrobend fill makes it a bit stiffer than an empty tube. If you do not want to bother with Cerrobend, you cannot use this method. Sand will not work.
Cerrobend is so easy to use, I can't see why you would avoid a simple method of making essentially perfect bends. Try it, you will like it.

WOB

 

[petertha]

Very interesting WOB, thanks for the pic. I just assumed it was the typical half circle groove die & friction slider block for outside. I've heard of Cerrobend but assumed it was more about even more difficult alloys. I'll have to get some. My friend told me he uses a hand held tool for tightening up decreasing stock bender radii but its more for final fit-up adjustment & is hard not to introduce the small segmented look. so I held off on that.

Getting back to the head washer / seal my plan was to use the copper based brake like on the exhaust pipes as-is. No particular reason other than look I suppose. But the aluminum versatube will be used on the induction side because it has more complicated bends & longer length. The copper is heavy although who knows if this engine will see air service. It was the aluminum (tube) directly on aluminum (head) & requirement of air tight seal that I thought a gasket washer or O-ring would be necessary. What did you do on that front?

So you made a right angle on your tube, can you describe how you did that? I think I could accomplish this on my forming plate using a second operation flat side press. The tube seems quite forgiving. But like I I was saying it was more about accommodating the resultant OD in the threaded hole. I'll experiment more.

 

[WOB]

The 3003 alum tube was softer than my heads which were 6061-T6. There was no need for a seal washer. The flare was swaged down onto the head recess bottom(made with a regular end mill) and the seal was gas tight. I used the alum tube on both intake and exhaust pipes with no problems. If you use copper for the exhaust pipes, you might need a gasket between the flare and the head recess, but I doubt it. A tiny leak on the exhaust pipe flare would not be a cause of trouble. It certainly won't affect the engine run. If you must have a washer, I would recommend 1/32" 3003 aluminum sheet.

As for bending the tube, with my little jig, all you have to do is hold the jig in a vise. Grab the tube ( Cerrobend filled) and bend it over the half round block with your hands. There is no spring back. You don't need anything else except maybe a cheater pipe ( tube?) It could not be more simple. It might work with the copper tube( it is not as malleable as alum) but it will be harder to bend. I have not tried it on anything but alum. Drop the bends into a pot of boiling water and all the Cerrobend will run out and there you have it. Don't over think this process. It is not that hard to do.

WOB

 

[petertha]

Getting closer. Received the taps last week. The 7/16-28 TPI is a better thread size. Grinding down a plug tap to make a very low bottoming tap also worked great.

Upon assembly the nut didn't quite bottom out the tube flange by itself but dropping the .020" thick washer made a seal. So I can either taper relieve the last thread that amount or use a similar dimension seal/gasket ring between the flange & counterbore face. I'll now fiddle around with a 90-deg flange to see what I can do & I have a line on some Cerrobend. A few more mockups on the valve cage & I am running out of excuses to make the heads for good.