Internal spoke nipples configuration and reliability?

[ausdb]

To all the wheelbuilding guru's I am after some help please if you don't mind.

I am interested in comments regarding the reliability of two different scenarios for constructing a wheel with a medium section alloy rim (say 27-28mm deep) and internal nipples:
Scenario A: Using an unslotted standard style nipple that is inverted, and a spoke length where the threaded section of the spoke bears against the actual wall of the rim.
Scenario B: Using a nipple that has a counterbore (doesnt matter if hex or standard) so that where the spoke passes through the wall of the rim it is unthreaded and the counterbore allows the nipple to make up tight without bottoming out.

If my descriptions above are no good I hope my sketch illustrates what I am asking about.
Internal_spoke_nipple by ausdb, on Flickr

To me scenario A seems like asking for trouble as whilst the spoke is loaded in tension if there is any flex in the rim this will create a shear loading on the spoke and the threaded section of the spoke where it goes through the rim is a stress raiser crying out "Snap Here Please". Scenario B will still see the same loadings but moving the threaded section of spoke removes the obvious stress raiser, am I correct in my thinking?

For a bit of background I weigh 180lb and have been riding a set of homebuilt Kinlin XR300's 20F, 24R (2X) made with some ACI bladed spoke and standard non internal nipples and not had troubles with spoke breakage, I also have a set of low end Shimano factory builts 20F, 24R (2X) with standard type nipples and have not been breaking spokes.
The problem is I have just bought a new bike with set of Roval Fusee EL28's which are 20F, 24R (DS= 16@2X, NDS=8 rad) and and in less than 1000km been suffering multiple broken spokes on the rear drive side. I think that the reason for this happening is that the wheel is built as in Scenario A and where the spokes keep breaking is the point where the threads of the spoke pass through the rim. Every time it has happened there has been about 1mm of thread attached to the inverted nipple and there is always about one thread left on the end of the broken spoke.

Thanks for your help.

Mods please move this to the wheelbuilding topic if you don't think it deserves its own space.

[youngs_modulus]

Hi Ausdb,

I don't have an answer for you yet--I'm still thinking about your question. However, I wanted to acknowledge the effort you put into making yourself understood. Rather than madly typing at your computer with a half-formed idea and no clear question, you've clearly given this a lot of thought; also, I found your diagram helpful.

So my response is essentially: "Good question, well-asked. I'm thinking about it."

Cheers,

Jason

[rruff]

The problem is I have just bought a new bike with set of Roval Fusee EL28's which are 20F, 24R (DS= 16@2X, NDS=8 rad) and and in less than 1000km been suffering multiple broken spokes on the rear drive side. I think that the reason for this happening is that the wheel is built as in Scenario A and where the spokes keep breaking is the point where the threads of the spoke pass through the rim. Every time it has happened there has been about 1mm of thread attached to the inverted nipple and there is always about one thread left on the end of the broken spoke.

That failure would be consistent with a bending load at that point in the spoke, and a lack of stress relieving. I don't think the rim is going to cause a shear load... even if the rim flexes, the nipple isn't going to be sliding around... it will just pivot slightly. It's a bit odd because nipple failures tend to happen more on the NDS because the angle is greater, and lower tension can be a contributing factor also, though with triplet lacing the tension would be nearly the same.

At this point I'd vote for a bad build (or defective spokes) being the cause, rather than the nipple design.

[em3]

Just to be clear and to state the obvious (and I mean no offense), but you are aware that rims intended for internal nipples are different then those that are intended for standard external nipples? Internal nipple rims are reinforced at the area of spoke bed and often are custom bore internally to accept a specific nipple (e.g. Campy locknut nipple)...boring the internal side of rim to match perfectly with the associated nipple avoids the scenarios you are describing above. Also, the holes on an internal nipple rim are usually smaller, which both add strength to the nipple hole in order to support the internal nipple and also help seat the spoke firmly in the hole. I have never heard, nor would I recommend, using a non-specific internal nipple rim with internal nipples, especially using an inverted standard nipple as an internal nipple, as you describe in Scenario A.
Good luck,
EM3
ps - all this talk about NIPPLES and your graphic sketches are making me tingly

[rruff]

I'm not sure what you are referring to, because the Edge rims are the only ones I know of that have a molded spot for the nipple. The others have a smooth surface all around the nipple-bed of the rim. Plus *external* nipples require more rim reinforcement because of the bigger holes and small contact area. He says the original nipples are not slotted, so those are likely made for internal use, even though they sort of look like a standard nipple turned upside down.

[hockinsk]

I don't see why you would want the thread showing in A. Just buy spokes long enough so all the thread is inside the inverted nipple. That's what I do anyway. Never had any problems using normal nipples upside down. In a way they lock nicely into the rim bed providing a sort of mechanical lock.

[ausdb]

Thanks for the comments so far

@youngs_modulus, cheers I try to work on the theory that in most situations it is better to think before engaging the mouth or keyboard.

@rruff, Thanks for clarifying my error regarding the shear loading and it actually being a bending load, I have some more questions about this below if you don't mind having a look.

@the other guys, sorry to be blunt but( that was not very polite) please read my original post. I am not considering in any way building a set of wheels in this manner I am trying to work out why I keep breaking spokes on a wheel that has the same spoke count as other wheelsets I own but are not breaking spokes on. The manner in which this wheel was constructed/designed by someone else seems to be the key to the issue (to me at least).

In regard to rruffs the comment about it being a bending load and the root cause being a build or faulty spoke issue I hope the following provides some more information.
I did a quick autopsy on the wheel before I took it back to the LBS and have posted the photo's here: Roval Wheel I took the step of doing this as the last time a spoke broke the LBS had binned the old one and the nipple befere I got a chance to see them.
The holes in the rims show clearly show signs of the spoke threads wearing against them and excuse the poor macro photos but when you look at the inside of the rim from the outside there is a distinct wear patch from the inverted nipple at only one point in the nipple seat area. I am wondering if this has something to do with the bending load being seen on the spokes where they pass through the rim wall? There is also no sign of any lubricant residue in the nipple seat area.

In regard to the spokes Specialized's website says they are DT New Aero and are 3.0 x 1.5mm bladed straight pulls. I have measured them with a vernier and come up with 3.0mm x 1.3mm, what I find strange is that the tech page on DT's website lists the New Aero as 3.3mm x 1.1mm ?????? so I am not really sure what spokes they are.

The last time the weheel was repaired by the LBS I checked the spoke tensions with a TM-1, on the Drive side the TM-1 readings were all between 14-14.5-15 and on the non drive side all were between 10-11 on the TM-1. I am not sure how to convert these to actual tension readings as the this spoke size isn't in the tables for the TM-1 and I can't find seem to find a formula to work it out (I do recall seeing something on the web but not anymore). I did a comparison with the front wheel and sampled a few spokes on it with the TM-1 which showed readings of between 12-13.

[hockinsk]

The problem is clearly the use of too short spokes on the Rovals. All stress is being concentrated on a threaded part of the spoke. Add to this inverted nipples without the un-threaded recess and it's asking to be prone to breaking spokes on the drive side.


The image below is basically a big mistake even using nipples the right way around. Either the nipple will shear when tensioning the wheel, or the spoke has a high chance of breaking where the thread exits the nipple. Sounds like Roval/Whelbuilder has got it about as wrong as you can get as they are doing this with inverted nipples too. There should really be 1-2mm of thread inside the nipple no mater what way around you are using it. As i said, i've built wheels with inverted hidden nipples and raced for two years on them without a single breakage. Deep carbon rims though, so much less movement in the rim and spokes compared to a a shallow alloy rim.

[rruff]

The problem is clearly the use of too short spokes on the Rovals.

As your photo states, the problem with short spokes is that the *nipple* can break. That isn't the issue here. Regardless of the spoke length, some threaded part of the spoke will be sticking past the nipple threads. If there is a significant bending force on the spoke where it exits the nipple, and the spoke was not adequately stress relieved, then problems are likely.

On second thought, it's possible that longer spokes could solve the issue if the spoke is pressing hard against the rim. In that case the rim would "support" the threaded portion of the spoke and the higher bending stress would then occur on the unthreaded part where it exits the rim.

[ausdb]

partially snipped for context
That isn't the issue here. Regardless of the spoke length, some threaded part of the spoke will be sticking past the nipple threads. If there is a significant bending force on the spoke where it exits the nipple, and the spoke was not adequately stress relieved, then problems are likely.

On second thought, it's possible that longer spokes could solve the issue if the spoke is pressing hard against the rim. In that case the rim would "support" the threaded portion of the spoke and the higher bending stress would then occur on the unthreaded part where it exits the rim.

I guess your last comment is what prompted me to draw the two sketchs in the beginning, A = how they did it and B = How I thought they would do it

What I don't understand is where is the bending force coming from, which is doing the damage is it just the nipple / spoke interface or something else?

The spokes are straight pulls so there shouldn't be any issues with the set of the spoke head in the hub, there also doesn't seem to be any windup in the spokes at all. I did not get a lot of tings and pings when I was squeezing pairs of spokes after the last time it was repaired so I am not convinced about stress relief being the cause.

The thread marks in the spoke holes of the rim show that there is movement at this junction, the single contact patch inside the rim from the spoke nipple suggests that the spoke is not enetring the rim at this point entirely perpendicular so would have to be applying some sort of bending force at this point? (please correct me if I am wrong)

I am by no means a powerhouse and generally tend to sit and spin my way up hills (reality = huff & puff) rather than stand up and stomp them so I don't think it is due to any outrageous power outputs on my behalf.

At the moment the wheel went back to the LBS who have passed it on to Specialized so I wait with baited breath, some forum searching suggests to me that this is not an isolated issue with this model.

[hockinsk]

As your photo states, the problem with short spokes is that the *nipple* can break. That isn't the issue here. Regardless of the spoke length, some threaded part of the spoke will be sticking past the nipple threads. If there is a significant bending force on the spoke where it exits the nipple, and the spoke was not adequately stress relieved, then problems are likely.

I wish people would read all of a post without taking the first few words out of context only to go on and say exactly the same thing as the post they quoted. I did go on to say:

There should really be 1-2mm of thread inside the nipple no mater what way around you are using it

Either the nipple will shear when tensioning the wheel, or the spoke has a high chance of breaking where the thread exits the nipple

[Seamus]

ausdb,

Nice description of the problem - I wish I could explain as clearly the answer though I would like to add some 'food for thought'.

The important aspect in this issue are the spoke threads. Threads are formed by a rolling process and generally have a failrly sharp profile. When you have a sharp profile at the root of the thread (meaning also a reduced spoke cross section) you have a stress raiser. Also the first full thread takes the highest share of the axial load and the following threads take less of the load.

I think the difference is the threads (profile and root diameter) between parts failing or not in your scenarios...
Seamus

[rruff]

There should really be 1-2mm of thread inside the nipple no mater what way around you are using it

Nope... it is impossible with the nipple he showed... plus typically the unthreaded part of the nipple (either way around) provides zero support for the spoke.

[rruff]

What I don't understand is where is the bending force coming from, which is doing the damage is it just the nipple / spoke interface or something else?

Basically the nipple cannot (or won't) rotate inside the rim so that it points directly to the hub flange... so there will be a bend in the spoke. The nipple will tend to be perpendicular to the rim bed. This puts high stress on the outer part of the bend, and then with cyclic loading you'll get fatigue.

The threaded part of the spoke is weak, so you want to avoid any added stress to that area. When I build a wheel like this, I bend the spoke above the threaded part so that it points in the right direction, then stress relieve.

[ausdb]

Just updating this topic let you know what the final resolution, many thanks to rruff for helping me understand the mechanism of why the failure was occurring.
After the second lot of broken spokes on the original Fusee EL28's Specialized supplied a new back wheel, that lasted for 500km and broke DS spokes again too.
Specialized then took the Fusee's back and replaced them with a set of the Rapide SL 45's which have been going for over 1000km now with no issues. They are somewhat wasted on me at my fitness level but they look nice on the bike and don't really weigh anymore than the old ones.