GMM

Also posted in ‘mechanics forum’.

Does anyone know what is considered an acceptable amount of drift, in mm’s, when gauging a carbon bike’s rear triangle?
Just returned a Felt carbon – rear wheel was tracking 3-5mm, left.
The LBS' owner says that's acceptable, his mechanic says otherwise.

unterhausen

By throwing out a number, you really haven't told us anything. How are you measuring drift? String test? Clearance from stays? CMM? What is your reference, are you using the same reference as Felt?

I have heard that a large American manufacturer says that 5mm is straight enough. Most steel builders try to stay a lot closer to straight than that, but there are plenty of pretzels out there that people are riding happily.

mudboy

Carbon and aluminum should be perfect. Period. Neither can be (well, should be, in the case of aluminum) cold set.

unterhausen

What does perfect mean in this instance? We don't even know what dimension is off by that much.

The problem with an open mold carbon bike is that there are no reference surfaces that you can trust, particularly assuming bb30. I'll bet there are very few carbon frames that don't have some dimension that is at least 3mm off. 3mm is pretty close for production work.

GMM

The front of the rear wheel(at the rim) is running about 4mm towards the left chainstay. The caliper mounting hole is even offset from the wheels center.

Used two methods to measure:
The shop used a long Park guage(ran from the headtube, across the seat tube and measured the distance at the drops)
Then used a straight edge ruler across the BB-shell, overhung it over the rim(above the chainsay) then measured the distance from the rims edge to the straight edge ruler. There's about a 4mm difference between the two sides. It's very noticeable.

Is that considered acceptable by carbon's standards???
At $1600, frame alignment wasn't even a considering factor during purchase.
Is such common among Felt?

unterhausen

None of those measurements alone really prove a problem, but the ensemble of measurements is pretty damning. Like I implied in my first post, Felt may consider this acceptable. They may look at it as being 2mm out on each side. Is it bb30?

I am curious, did you notice this because of anything in the ride, or just visually?

GMM

It's a standard bb, and very notible.
If Felt, or anyone else in the industry, considers such acceptable... that's pathetic. We've owned all kinds of framesets and never had an alignment issue, even with a cheapo $600 bike. In my opinion, if the tire aint centered inbetween the chainstays AND caliper bridge... the frame ain't worth the time. You're better off going custom.
Would you agree?
Look at it this way; what's the point in spending countless hours truing a wheel if the frames alignment sucks???

unterhausen

sounds like you did the right thing in returning it. I am not sure what else to tell you, from what I've heard, carbon production bikes are all over the map alignment-wise. So buying something else might not get you anywhere.

Andrew R Stewart 

GMM- Why true a wheel to a few thousand thousandths of an inch only to mount a tire that's off by a few mms? (That was said, more or less, by the head of Santana about 30 years ago as a reason he never used sew ups..) Another words, it's all relative. And how the related dimensions/alignments line up are the real ride effect. As an example one could have a frame that was off center but a wheel that was dished to compensate. the tire contact point with the ground would be in line with the BB and ft wheel but it would look all cock eyed to the rider in it's draft.

Eric is right to question how the stated dimensions were measured. My coworker thought that the carbon frame he was building up was off before he learned that the seat tube was asymmetrical by design. Rendering the string test invalid. A best test would have the wheel's alignment measured (think of two pairs of very long straight edges running along either side of the rims parallel to the ground front to back, one set a few inches off the ground and the other set a few inches below the rim's tops). Then all frame issues between the wheels would not matter excepting those that misaligned the wheels ONLY.

How "straight" a carbon frame (or fork) is is a debate. How straight it needs to be before a rider will notice it when riding is another debate. (I've had riders say one thing about their high buck bike then change their tune after showing them the alignment issues that existed). As a guy who works with steel I can get the frame/fork to measure real straight using traditional builder alignment methods but still have any one rider not be able to ride no handed without leaning a bit to one side or the other side.

Did you ride this bike?

I will agree that it seems that the frame was not well aligned. Certainly a big enough question in your mind to not have it satisfy you. This lack of confidence is enough a reason to reject it.

So many assume that a molded frame is as straight as the mold is... Andy.

GMM

I beleive I made the right 'return it' decision, too. If that's what carbon's all about, it ain't my thing.

Any opinion on Marinoni steel framesets? $1075 for custon geometry is very attractive. That's Columbus Spirit traingle with Zona stays.

tuz

I haven't ridden their tigged frames but a friends loves his. FWIW I've dealt with them quite a few times, great people.

GMM

Well, I thought I could return the frameset for a refund, but it turns out that one of Felt's rep's insists that 3mm out is acceptable, AND, if I return the frameset for them to jig, and it's 3mm or less, they will not refund, exchange or pick up the shipping fee. How's that for service!

So, the question is:
1. What measuring method can we use to verify alignment?
We did a string method, (bike in stand, wheels in place, place string around headset spacer then run each end back towards the drops and lace in-between each sides of each rear wheels axle nuts and dropouts contact points). It's a straight unobstructed route, both sides.
Then we measured the space from the seat tube to each string (left and right side) and sure enough, the left side string runs about 3mm more outboard then the right side.
Next, we installed the crankset and measured the gap between each crankarm and chainstay and again, the left runs about 3mm more gap then the right.

The problem is though is that it's difficult to know just how accurate that particular string method is when used with a carbon bike's carbon seat tube.

Another option: Remove the crankset and cups, place a long straight edge ruler on bb shell, point straight edge vertically past each string, then measure the distance from the string to the straight edge, both sides. Would that be considered an accurate method, or is there a better way?

We need definite proof that the frameset exceeds Felt's 3mm rule in order to refund or at least, exchange. (The LBS doesn't have a builders frame jig)

The gripes I’ve got are:
1) How closely the rear tyre tracks to the left chainstay – guessing about 3mm (I’ll feeler gauge it next) but it’s so close that I can actually make tyre/chainstay contact by lightly squeezing wheel and chainstay.
2) That the caliper’s mounting hole is about 2mm off the tyres center.

Does this whole alignment issue seem like a legitimate gripe, or, is such really nothing to be concerned about?
I’ll also ask the LBS to ask Felt’s rep to indicate what exactly does the 3mm rule include, and how do we measure such, just the same. And I'll finish buiding it up and do a test ride, too.

Andrew R Stewart 

GMM- I think you need to get the frame to a person (builder) that has a proper flat surface and the tools to check alignment who will documemnt the process with both photos and numbers. The process would go something like this. The frame (and fork as it's part of the whole) is stripped of all parts. the frame is suspended on a through rod, with cones at each end of the head tube (this way only the internal reamed bore is used, not the outside surface of the HT), and the frame is supported parallel to the surface with a jack stand under the seat tube center. A rear axle dummy is placed in the drop outs and it's center is used to insure the frame is parallel to the surface. This dummy axle is checked for squareness with the surface (alternitely a good wheel could also be used here). The bb shell is checked for being axle centered height off the surface, as is also the rear brake hole and the seat post entry of the ST. Now you know whether the frame is in line front to rear or twisted, HT WRT ST and rear axle. The shell could stll be crooked but mounting a BB and crank (with old fashined tapered square and cup and cone design) both ways and measuring each arm to stay BOTH ways the BB axle is installed will tell you aout the shell's alignment (relying on the crank manufacturer to have symetrical arms after mounting is asking a lot). The fork is clamped by the steerer (if not tapered) or the dummy axle and the other end is checked for being centered/parallel to the surface. Twist is simply two straight edges across the axle and the top of the blades.

You can't rely on the out sides of a carbon frame to be symetrical to the frame's centerline. You can't rely on the edges of the Bb shell to be square (as in being faced with a cutting tool). You can't rely on any one "tube" being straight by design of the curves being symetrical. Only the insides or contact points are needing to be correct. This type of alignment is very hard to do without proper tools and understanding. Without dial or height gauge readouts any viewed discrepincies will be subjective and not "real".

I am not surrprised that Felt is hesitant to work with a consumer on this. How they take a pro's report will be interesting. That they allow 3x the common amount of missalignment (that many pro builders use as their tollorance) is telling to both the process of building with carbon (in a cost concerned production factory) and their company policies. Andy.

reddog3

Absolutely! We spend a lot of money on frames/bicycles, and have lots of pride on how they look, and how others view it. If it doesn't look right, it ain't right!

You might not notice anything in the ride quality, and it may well be within Felt specs, but it will always be apparent that the wheel isn't sitting right in the frame. You know it- others will see it. One way or the other you'll pay for this everytime you drag it out for a ride.

I'd bet anything that the LBS owner, or the Felt rep would not accept this of their personal ride, and you shouldn't be expected to either. I don't care if it is within specs. That should be the deciding factor for a return or a replacement. Were I Felt I'd want my stuff to look right. If it doesn't it's outta here!

Those of us here who build our own frames, be it for hobby or livelyhood, are not allowed these kinds of discrepancies. It reflects on our pride and skills as a builders. I'd like to think Felt has some pride as well, and cares about their reputations, even if it within their "loose" specs.

CF construction may be prone to take a "set" as it cures, but if it doesn't set right- it ain't right! Ask the Felt rep if he'd accept this, and if his personal ride suffers from the same problem. If the answer is no... that's reason enough for an adjustment.

GMM

Thanks guys, I completely agree! I'll ask the rep myself if he'd be willing to ride it day in and day out. I'm guessing not!

And, hopefully we can find a local framebuilder who can properly jig this thing in order to get the 'true' proof.

Thanks Again

Scooper

Where is "local"? Someone here might be abe to put it on an alignment table and measure it for you.

Dave Kirk 

If you walked into my shop here is what I'd do as a quickie check-

* put a rear wheel without a tire in the frame. Be sure it's seated as it should be and clamp it tight.

* turn the bike upside down with the BB up.

* if there are cables running under the bb shell do what you need to do to get them out of the way.

* put a piece of masking tape on the bottom of the BB shell.

* here's the thing that is hard to describe but easy to do - take a long straight edge and lay in wide side down on the underside of the BB shell and then touch the edge of the straight edge up against the rear wheel in two places. So now the ruler is flat against the shell and edge-on to driveside of the rim. It should be nice and stable like this as it has three contact points - just make sure the ruler isn't hitting the spokes - you may need to move the ruler further away from the hub if it's hitting spokes.

* with the ruler in place use a marker and make a mark on the tape on the bottom of the shell on the edge of the ruler. You have in effect projected the rim edge forward onto the BB. Try it again and see if you get the same results. If you don't, figure out why - you should get the same thing every time.

* now take the wheel out and flip it around so it's in the frame backward. Be sure it's seated properly and tighten.

* now repeat the ruler trick on the other side and make another mark on the shell. Double check your work.

* now measure how far from the edge of the shell each of your marks is. If they aren't the same then the rear wheel isn't pointed straight ahead. If they are the same you are good to go.


Measuring the distance from the rim/tire to a chainstay is not a good way to determine proper alignment. It could be all kinds of things that would cause this. But if you do the above it will tell you if the wheel is pointed forward or not.

Give it a try and see what you get and report back. It would be interesting to see what you get.

dave

Andrew R Stewart 

Dave's quick rear wheel check is some of what i was suggesting in my first post to this thread. My method of two straight edges along side of both rims would involve the fork/ft wheel with the rear. Not sure where one would get two , almost, 6 foot straight edges that could be sandwiching the rims... maybe 8020. Also both methods dosen't really consider the wheels' vertical alignment (not that the vertical controls the bike''s tracking much).

Dave- If I understand your method, it won't take into account the BB shell being offset. Might making two marks on the tape, from each side of the rim, then flip the rim and remark both sides of the rim be a better indicator of the wheel's alignment WRT the bike's centerline? If the wheel was well centered then the marks from the drive side of the straight edge, one each with the wheel installed both ways, be on top of each other. As the lines fron the nondrive side of the straight edge would also be on top of each other. In other words you'd have two pairs of lines and if the frame was not aligned well the two pairs of lines would be a little bit off set from each other. Andy.

Dave Kirk 

I think that could work. For rough checking I do as prescribed and it's fast and simple and it rules out issues with the rear wheel not being dished properly. There are a lot of 'bad' frames out there that are just fine and the real issue is bad wheel dish.

Dave

GMM

Hey Dave, did it just to note and here are the numbers:

BB shell width = 68mm
Rear wheels rim offset from bb-shells center: Right side = 7.5mm. Left side = 11mm. Which translates to a 3.5mm difference.

Also feeler gauged the distance(gap) from the tire(25c) to the chainstays(just to note) and they are: Left side = 1.85mm. Right side = 5.08mm. Difference of 3.23mm

Seems like no matter how we measure things (string method, ruler on bb shell to rim, crank arm to chain stays, or bb shell center/rim with tape) that 3mm difference is clearly evident.

Dave Kirk 

Very good - now you know were you stand and know for sure that it's not right. I hope they take care of you - they should.

dave

Canaboo

I hope some kid in Taiwan isn't going to get in trouble for this.