Trying to fix brake caliper banjo leak, desperation setting in

[Flojo]

Originally Posted by jwakil

The seal for a banjo joint is made at the flat mating surfaces, not the thread. That won’t work. Here are a couple of more pics of the surface and banjo, and my method for sanding the surface.

you sanded the surface on the caliper. so far so good.
what about the surface of the banjo?

[jwakil]

Originally Posted by Flojo

you sanded the surface on the caliper. so far so good.
what about the surface of the banjo?

That surface is fine. I have also flipped the banjo and doesn’t impact. I can also visually see where the seepage starts.

[Flojo]

Ok. then -other than using a sealant- I'm clueless as you are.

[jlex]

Proper torque?

[75 911s]

Originally Posted by jwakil

There is no way I can get that surface any better.

It could be machined off square. Take the caliper apart, and machine the surface flat again. I wouldn't trust sanding it down with a hand held drill.

[ryans65]

Can you screw the banjo bolt in a couple threads then measure with a caliper the distance between the head of the bolt and the sealing surface on the caliper in a few areas? I have a feeling the head of the bolt is off square from the sealing surface on the caliper. Its the only explanation that makes sense at this point aside from some small crack or defect in the sealing surface itself that cannot be seen with the naked eye.

[jlex]

Is that bolt bottoming out before it's screwed on all the way? Maybe the material you shaved off was just enough to disallow the faces to meet. Put the thing on without a washer and see if it tightens up to rule out the bolt bottoming out. May also be able to see how the surfaces are mating with the washer out of the way.

[bpu699]

You sure you don’t just have a hairline crack in the banjo fitting or bolt? These things just aren’t that hard to seal... did you over tighten/pancake the banjo fitting? You could also try composite washers that aren’t copper, they smush way more, though I don’t know their life span or if that’s safe...

[ClickClickBoom]

Originally Posted by jwakil

Well, I'm down to my last two resorts. I was very optimistic about the Napa washers, as they were exact size ID of the Porsche ones, but were thicker and larger OD. Tried two different ones, with and without annealing, and dam small seepage could be seen after about a minute under brake pressure. And yes, it is very obvious it is the caliper to banjo interface and specifically the caliper to washer interface. I have that surface smooth as silk. I cannot believe this is happening. It's almost like a curse at this point. There is no way I can get that surface any better. So last two resorts to follow:
1. Permatex sealant suggested by Rawknees Turbo
2. Replumb the lines so that I can put the banjo on the other inlet on the outside half of caliper. (least desirable).

Think about the problem. You have determined that the problem area is definitively between the caliper body and banjo assembly. If so unless there is a micro crack opening up under pressure, or the mating surfaces aren't parallel the only other possibility is the washer.
Unless the micro crack extends beyond the washer, doubtful you wouldn't see it with all the work you have done in that area.
Crush washers are designed to compensate for surface irregularities as well as sealing.
My bet is during the "rework" the mating surface became less than perpendicular. If your washers aren't soft enough they cannot deform and take up the irregularities.
Another option is aluminum washers to determine if it is a washer problem.
Also it may be seen as obvious, when you are quenching the copper washer, are you heating to a dull red heat, then quenching in cold water? If not the hardening of the material is not normalized.

[ryans65]

Heat it until it is glowing ****** hot then quench or simply let it cool. Doesn't matter. Now drop it on the floor and listen, if it has a metallic "ring" sound then you didn't do it right. If it hits the floor with a dull tap then its good. If there is some sort of surface anomaly larger than what the annealed copper washer can't seal then you need to seek professional help. PMB was mentioned above.

[jwakil]

Ok update. I tried the permatex sealant. By the way, $30 for a tiny bottle, ouch! Even left it for a few days to make sure sealant was properly cured before connecting the input line with fluid. While I had high hopes, let’s just say it failed immediately. Worthless in this application. That is a thread sealant and on flat surfaces it’s useless.
So finally went to the last option, which was to connect the banjo to the port on the outer half of the caliper. Then I replaced the incoming line with a custom made one which could reach the outside. It’s an ugly solution, but so far it’s working. It basically moved the banjo mating surface to a new unblemished one. So it proves that the caliper surface was messed up. It’s unbelievable that after all the sanding and insuring the surface was flat and perpendicular to the port (that was the whole purpose of the special tool I created) it still wasn’t good enough. The surface was mirror smooth, and I took great care to ensure that I was always grinding perpendicular to the axis of the port. Something was still not right and still cannot believe it. And, no I wasn’t bottoming the banjo bolt. I tried so many washers, thicker thinner, etc. my understanding is copper is the softest metal not aluminum. What would be a softer metal washer than copper? Anyway, at this stage I’m happy to just have a working solution. I don’t want to touch it at this point. At some point I’m thinking I might get rid of the banjo altogether by getting a T-junction and splitting the input line in two before the caliper, then I can bring a flared line to each half separately. It will be a little cleaner. Anyway, I don’t have a picture of my current solution, but first time I take the wheel off again I’ll take one in case you can’t picture what I’ve done.

[Rawknees'Turbo]

^^^

I'm really sorry that my suggestion didn't work out (surprised, too), and cost you $30 and some wasted time - apologies.

[ClickClickBoom]

Did you anneal your copper washers?
Pure aluminum is between 2.5-3.0 on the Mohs scale, pure copper is 3.0.

[jwakil]

Originally Posted by ClickClickBoom

Did you anneal your copper washers?
Pure aluminum is between 2.5-3.0 on the Mohs scale, pure copper is 3.0.

yes, I heated on my gas stove flame until pretty hot. Didn't seem to make much difference. I also figure Porsche washers would not require annealing by the user.

[jpnovak]

Originally Posted by jwakil

yes, I heated on my gas stove flame until pretty hot. Didn't seem to make much difference. I also figure Porsche washers would not require annealing by the user.

YOu have to heat them hot and then dunk in cold water to anneal. Letting them cool normally will sometimes increase the hardness.

If you drop them on the floor they should land with a thud and almost not bounce. Any "ping" sound and bouncing means they are too hard.

[ryans65]

We've already covered this and no, either quenching with water or letting them cool to ambient temp without water won't affect the hardness of copper like it will with ferrous metals. Just get it glowing hot then cool it however you please.

Originally Posted by jpnovak

YOu have to heat them hot and then dunk in cold water to anneal. Letting them cool normally will sometimes increase the hardness.

If you drop them on the floor they should land with a thud and almost not bounce. Any "ping" sound and bouncing means othey are too hard.