Plugging Hole in Pressure vessel via Staked in Ball 2

[BrianRoach]

Not sure where to post this so I am putting it here.

I am looking for design advice for plugging a hole in a hydraulic brake master cylinder by staking in a steel ball.
The hole I am plugging is shown in the image below:

The cylinder is filled with brake fluid at pressures up to 150bar. The hole that has to be plugged is approximately 6.1mm. The hole diameter can be varied as needed, but I want to us a 6.35mm ball for the plug. The cylinder is made out of extruded aluminum (6061). Currently we are using an Advil high pressure plug to fill the hole, but for manufacturing reasons and cost we want to change to a staked ball. The staked in ball would look like this image:

I need to determine the diameter of the hole, the shape of the staking tool, the required staking depth, the staking force, etc. Is there a way to determine these parameters through calculation or is the only way via experimination in the lab?

 

[unclesyd]

Have you looked at the ballizing process parameters for sizing a hole. A lot of ball suppliers are also in the ballizing business and there have the know how to keep from ruining the hole or getting the ball stuck. You might buy a few balls and get a lot of information.

Things like if you are pushing a ball through a hole too slowly you generate waves in the wall of the hole, especially true of Al. In your case this would be good because it would tend to trap the ball. By selecting the right size ball and insertion speed you could possibly generated an extremely tight fit in the part.

 

[hydtools]

I tripped over this as I was doing a search. Izumi mentions ABS ball insertion as an application of their press. Programmed insertion force and depth control.

http://www.izumiinternational.com/janome/Janome_full_catalog.pdf

Page 11.

Ted

 

[BrianRoach]

First, somehow this string got too personal. That was not my intent. Sorry if I upset anyone. Second, everyone has provided lots of good info. Let me try to reply back to some of the comments posted. I have researched and found the equations for determining the force to press fit a cylinder (pin) into a hole and think I understand these equations, however I am not sure they apply to a ball pressed into a hole. As I understand it, the force to press in a pin is F = friction coefficient * contact area * contact pressure. The contact pressure is a function of the pin/hole diameters, the modulus & Poisson’s of the materials and the interference. If I assume the contact pressure equation is the same for a pin in a hole as for a ball in a hole (which it probably isn’t exactly the same), I still have a problem in calculating the contact area. For a cylinder it is pi * Length * Diameter. For a ball, the theoretical contact is a line so L = 0, therefore contact area = 0, therefore force = 0. So what I am trying to find is force equation for a ball press fit into a hole. MikeHalloran, you mention this is the Machinery’s Handbook, but I can’t find anything about a ball pressed into a hole. Could you give a page number or index entry as a reference? hydtools, you mention to look at press fit equations. That is where I found the equations I mentioned above for a pin in a hole. Do you know of a good reference that discusses a ball in a hole? That is the crux of my problem. I have found good examples of a pin in a hole, but nothing on a ball in a hole. If anyone knows where I can find this reference, it would help me out tremendously

As far as the ballizing, I had not considered that before. Although, as I understand ballizing it is not used to plug holes, rather it is used to resize holes and improve the surface finish. However, the companies that manufacture the balls and ballizing equipment might be a good resource of information for my problem. I will try to contact one of them to see if they can help out. Thanks for the idea.

 

[btrueblood]

Sorry to rake you over the coal, Mr. Roach, but we need to have all the legal disclaimers up front. You seem to be ready to spend a lot of development money and test time up-front for this project, and to understand the risks, which is what we were asking.

"For a ball, the theoretical contact is a line so L = 0, therefore contact area = 0, therefore force = 0."

No, look at Roark's Formulas for Stress and Strain for Hertzian contact. For your situation, the hole will expand, the ball will contract, and you will get a funny pressure distribution over a somewhat cylindrical section of the ball's surface. Not a simple problem, but if you dig into the references in Roarks, and go to the source (Timoshenko may be a good place to start) you may find some classical analysis that will help you predict better.

As far as the staking/peening operation on the backside of the ball, you are dealing with a plastic deformation process. Classical analysis is pretty limited, but Slater's Engineering Plasticity has some pretty good descriptions of the mechanics and some of the old-style solution methods (method of characteristics) that can give you a place to start. Past that, you will want to look into FEA codes that can handle material and geometric non-linearites (large plastic deformations), these were fairly cutting edge stuff 10-15 years ago, but are becoming somewhat common today. What is not common are material parameter handbooks to tell you what the various elastic/plastic material model coefficients should be for your particular material(s), nor far less what the likely 3-sigma distributions of those parameters may be; you will likely need to pay to have a reputable lab develop a large number of stress/strain curves, under various loading conditons, to get to where you ought to be.

At some point in the analysis, you may find that plastic staking from the backside of the ball may not reliably maintain a compressive load to hold the ball against the bottom of its hole, and the ball can in some cases slip backwards, however slightly, ruining the sealing line. The expansion plugs can have this problem as well, and (some plugs) use an elastomeric coating or other seal element (e.g. multiple ridges to create multiple seal lines) to help increase the friction coefficient and development of the seal line by plastic displacement of the elastomer. Maybe you could put a stepped hole for the ball to bottom against, and put an O-ring into the corner between teh ball and the step, would seem like a more reliable way to seal something than just pressing in a plug.

Sorry, that rambled a bit, but as somebody said, you are getting this for free.

 

[MikeHalloran]

MH covers the shaft in hub case, which is a starting point. I did not mean to imply that it covers the ball case; it doesn't.

As btb points out, you need the housing to behave elastically in order to retain some radial pressure at the sealing surface... and when you induce plastic behavior nearby, e.g. by staking, you may inadvertently relieve some of the elastic stress you need... and then it gets complicated.

How about an autogenous EB weld of a simple aluminum plug?





Mike Halloran
Pembroke Pines, FL, USA

 

[BrianRoach]

Sorry for my late reply, I was out of the office. The welding is and intresting idea that I had not considered. I discussed this approach with my manufacturing engineers and they don't want to try and tackle welding at this time. I think I will have to stick with ball staking for now. I have been doing extensive literature searches on the engineering theories for plugs pressed into holes and am hoping I can find something there that will give me some hints on how to proceed. Unfortunately most of the examples are for cylinder plugs (haven't found one for a ball yet). It will take me some time to try and modify the equations from a cylinder to a ball. I guess that's why they pay me the big bucks. (ha ha)

Thanks to everyone for all of the input. It helped to guide me in the right direction!

 

[EdDanzer]

Staking will be similar to riveting. You must yield material from the side of the hole into the center. The ball will be trying to yield this material back as pressure pushed it out. Using a cylindrical plug will be less costly and require less material to be deformed than a ball.

Ed Danzer

http://www.danzcoinc.com

http://www.dehyds.com

 

[debun]

Ah someone with the same problem that I have. I’m trying to plug a brass housing but my problem is more complicated because of galvanic corrosion so I’m thinking of using a brass ball or 316 which are drastically softer than say 440C or 52100. I also have to withstand 3250 PSI and seal against gas leakage. Now that’s exciting.
There is no way I’m going to solve it without testing and I guess you are in the same boat but looking for a starting point. Well I did some googling for ballizing, ball burnishing etc. and didn’t get too far. But I think there is a lot to be learned from those guys. If anyone comes across any good resources on bollixing or ball burnishing please let me know.
As for my starting point . I plan to use a press fit for 0.0005” to 0.002” interference. Since Aluminum has a lower modulus you can probably go higher.

 

[unclesyd]

Have you queried these people about your problem.

http://www.precisionballs.com/index.html

 

[israelkk]

Why are you trying to reinvent the wheel. Lee Plugs from the Lee Company are cheap and reliable and have proven for many years.

 

[debun]

Unclesyd: Those guys are bad, ass thanks.

http://www.precisionballs.com/Ball_Sizing.htm

Israelkk: I'll look into those Lee plugs but I work in corporate American on a product sold to Americans. I'm re-inventing the wheel for the same reason everyone is outsourcing to the country with the lowest labor costs. Must be nice to be in Aerospace

 

[btrueblood]

see United States Patent 6679953

Basically, they nitride the steel housing to cause the steel to expand, and better seal the ball after press fit. Perhaps a similar process (anodizing?) could work in your case.

Still think you'd be better off incorporating $0.0003 per unit of some elastomeric element under the ball to ensure a seal.

 

[debun]

Israelkk: I took at look at the lee plugs, and I quote from the website "It is slipped into a reamed counterbore in the product" at this point sounds like a backup plan if I can't get the ball to work. Thanks

 

[israelkk]

debun

This is what nice about the Lee Plug which is freely inserted into a reamed hole and then a coned small cylinder is pushed/hammered into the plug pressing it radially. The plug also has a series of round slots that to my opinion serve as a labyrinth to avoid leaks. So basically the plug is in contact over a cylinder while a ball will be in contact with the body only across a round circle/line.