Frame Alignment-Paul Brodie

[tnort...@gmail.com]

He posted a new video. So now I'm waiting to hear the feedback as I am a newbie!!

Tom

NE Ohio

[pa...@ellisbriggscycles.co.uk]

Aligning an old frame involves some best guesses but I would definitely not align a new frame that way.  Those methods don't take into account bow in the tubes.  Even the best tubes are not perfectly straight.  So if you rely on one reading in the middle of the tube for example then that maybe where the tube is slightly bowed and it will give a false reading.  Then you may make an adjustment where you pull the frame a long way out of alignment to satisfy your false reading.

Also there are a number of reasons why your frame may give a different reading when you flip it on the faceplate.  One of which would be how centrally the tubes are in relation to the centreline of the frame.  A downtube for example could be slightly off the centreline but the head tube and bottom bracket could still be inline. so then if you flip the frame the reading would be different.

In my experience though, most frames are not particularly well aligned.

I can see why many framebuilders choose those methods though, as they are quick.  They are necessarily accurate though.

Its the little amount of time spent on the rear dropouts that always puzzles me.

Paul Gibson

Ellis Briggs

[Guy Gadboit]

On Monday, 22 February 2021 at 00:26:20 UTC pa...@ellisbriggscycles.co.uk wrote:

> Aligning an old frame involves some best guesses but I would
> definitely not align a new frame that way.  Those methods don't take
> into account bow in the tubes.  Even the best tubes are not perfectly
> straight.

Exactly right. In fact the straightness tolerance of a Reynolds tube is
about 1.5mm "TIR" which I think means it's allowed to bow up or down a
maximum of 1.5mm over its entire length. I don't think they're usually
that much but that is the maximum allowed.

I guess a really fussy framebuilder should measure this before deciding
how to clock the tube in the frame. For the main triangle you'd want the
bows in the plane of the frame not going out to the sides.

> So if you rely on one reading in the middle of the tube for
> example then that maybe where the tube is slightly bowed and it will
> give a false reading.  Then you may make an adjustment where you pull
> the frame a long way out of alignment to satisfy your false reading.

I also think that if you correct a TIG welded frame like that with the
BB shell clamped down then if there wasn't a bow in the tubes to start
with, there probably is now.

I've heard that with a lugged frame the lug might actually bend and the
tube stay straight, which is what you want.

But I don't see how a curved ST or DT is necessarily helping matters
much with your frame alignment.

I think with a TIG frame you're better to align the HT, ST and axle. You
get the BB shell as square as humanly possible to the ST when you weld
it (careful fixturing, tacking and weld sequence) and then you have to
accept that it's not going anywhere after that. An error of 0.1 degrees
is not going to do any more harm than trying to correct it would.

If I had a table like that I would use it to measure the crap out of
everything, including the tube straightness. Then a lot of
head-scratching to decide what was actually wrong and what, if anything,
to attempt to correct, before breaking out the car jack.

My own experience is that the main triangle doesn't move much if you
fixture it right and weld it carefully. It doesn't have anywhere to go
and the mitres hold it in shape pretty well.

The rear triangle will always move around (usually it just closes up)
but so long as you do CS, align, CS bridge, align, SS, align, SS bridge,
align, it's not too hard to get it right.

> Also there are a number of reasons why your frame may give a different
> reading when you flip it on the faceplate.  One of which would be how
> centrally the tubes are in relation to the centreline of the frame.  A
> downtube for example could be slightly off the centreline but the head
> tube and bottom bracket could still be inline. so then if you flip the
> frame the reading would be different.

Yes good point. And you don't really care where the DT is so long as the
HT and ST are parallel and both on the centreline.

[Hahn Rossman]

I thought the most interesting part of the video was his "hooks" fixture to allow tweaking the frame that wasn't attached to the whipping post.

I also think that you should only be using the headtube as datum, and what the tubes are doing on the way there (bowing, going straight, etc) doesn't matter. I was also surprised that he didn't mention how small a defect on the bottom bracket shell translates to what seems like a huge error on the alignment table. In the downtube example, probably a 600+mm tube, that's a 20:1 magnification of that surface! 

Hahn Rossman

--
--
You received this message because you are subscribed to the Google
Groups "Framebuilders" group.
 
Searchable archives for this group can be found at http://groups.google.com/group/framebuilders (recent content) and http://search.bikelist.org (older content).
 
To post to this group, send email to frameb...@googlegroups.com
To unsubscribe from this group, send email to
framebuilder...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/framebuilders?hl=en

---
You received this message because you are subscribed to the Google Groups "Framebuilders" group.
To unsubscribe from this group and stop receiving emails from it, send an email to framebuilder...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/framebuilders/248223e7-8cc1-4e2d-b93d-acd8c40b1768n%40googlegroups.com.

[Brandon Ives]

Tom,
We’ve had some really good discussions in the past about this topic, so you might want to check the archives too.

That said I have some thoughts on alignment and Brodie’s video in particular. I totally agree with Paul that you don't need a surface table to make a well aligned frame. I say this as someone who aligned 5-7 frames a day for two years on a surface table 20 years ago. A table is fast and easy and gives a pretty good result, but the problems that Paul points out are completely correct IMHO. There are plenty of reasons you may want to own a good surface table, but correcting a post-build alignment should be pretty far down the list.

You need to start by thinking about what “perfect” alignment actually is and does. Ideally, you have the wheels and seatpost on a two-dimensional plane that bisects the rider. I’m going to get a lot of hell for this next part. A bottom bracket shell does not need to be all that centered or straight. It is a good central datum for alignment on a surface plate, but that’s about it. Many cranks and pedals are not perfectly aligned either to say nothing about human legs. Beyond that it’s a dynamic structure and flexes out of the perfect plane with each pedal stroke. This is why I think basing the alignment on the headtube or rear dropouts is better option if you can.

Looking at the frame Paul Brodie used in the video it shows exactly of how little importance perfect alignment means for the ride of a built up bike. That frame was obviously hammered and served for many years and was also out of alignment. I’ve checked dozens of classic race frames and many modern ones on tables and very rarely are they anywhere close to “perfect.” If you ride a bunch of bikes and also check their alignments you’ll find that alignment and ride quality is not closely correlated. All that said, the stiffer the frame the more noticeable poor alignment will be during a ride.

In the end if you like aligning on a table with micrometers great, but if you like the Taylor brothers holding it up to a window frame and eyeball for frame twist cool too. If the frame rides straight and tires don’t rub you're good in my book. Having a well aligned frame is important, but proper design and construction is more important to that than any alignment technique.

Best,
Brandon Ives
Springfield, MO

[Mark Stonich]

And then there are tales of bikes developing a speed wobble after a “professional alignment”. 

A gent in our human powered vehicle club built a long wheelbase recumbent and said he liked the way it handled. After he’d ridden it a few years I followed him through a puddle and saw that he left 2 tracks a good 4” apart. He was 6’8” tall, so none of us could test ride it. ‘Bents are generally not ridable “no hands”, but if he’d tried I expect it would not have gone well.

I had to replace the BB on my son’s race bike and was off by an embarrassing amount. He said it was fine and rode it another 20 years.

Mike Burroughs built a few MTBs with a single sided suspension fork. The fork tube was centered in the HT and the wheel was offset to the right. Some guy won several important races on them.

[Guy Gadboit]

A long time ago I saved up my pocket money to buy some MA2 rims to
upgrade the shiny chromed steel ones on my 12-speed (which in those days
meant 2x6) Falcon.

There was no www back then and certainly no sheldonbrown.com. The only
source of tech advice was what little you could obtain at great cost
from the local misinformed and curmudgeonly bike shops.

I laced them up all right, copying the pattern from the other wheel that
I hadn't taken apart yet. But it never occurred to me rear wheels had a
thing called "dish", so I just centered both rims between the flanges.

Five straight hours of "truing" later and it was just about round and
went back in the bike. It looked a bit funny but British bikes of that
era had diagonal horizontal dropouts, possibly because the frames
weren't aligned too well.

There was enough wiggle room to centre the wheel in both sets of stays,

so that's what I did, which meant it was tilting to the right of vertical

and also steering right.

The bike rode absolutely fine and I kept it for about the next 10 years
until it got hit by a car.