filing/mitering from archives

i'll likely be doing this type of post a lot.  it's the best way i can think of to cull valuable info from the email list archives.  my procedure is to search for something i'm interested, in this case mitering with files, then read through the results until i find something interesting from someone who seems authoritative, or even newbies describing their first experiences.  i copy the relevant text into an email message, which i send to a special address so it gets posted here.
in essence, i'm taking the 1st step towards making a comprehensive FAQ of the list, although that's a job i realized i'm not up for after reading through just a small portion of the archives.
these posts are not meant to be concise or highly readable, but they're almost pure data, and they'll be my guide for most procedures.  this is my paterek manual.
intentionally, there's a lot of opinions here, some conflicting, some agreeing.  my goal is to get together a survey of the info in the archives, not to present anything comprehensive.
From: "Anvil Bikeworks" <ojv(AT)>
Subject: RE: [Frame] Diamond Files
Date: Tue, 16 Mar 2004 01:04:46 -0700

As I said yesterday, the answer is to replace your files more often if you're working with hard tubes.  Using a diamond file to miter a tube is like what, help me think of a witty metaphor here...using an organically grown q-tip to wear a hole in a tube for a water bottle boss?  Something along those lines.  Anyway, you're going to be there a loooong time and that diamond file is going to cost you royal.  So you'll end up spending 5x more money for the privilege of having to spend 5x more time trying to wear a miter into a tube.  There are carbide files on the market but they're a) too coarse for mitering tubes, b) not available in the half round shapes we use, and c) too expensive.

I've seen a lot of reference to the Simonds Maxi Sharp files.  The Maxi Sharps are fine files, but they're just black oxide coated, which while I agree makes them look cool, actually does nothing for their performance other than making them slightly more resistant to rust.  As long as you don't have to pay more for them than you do regular files, they're fine, but
they're not going to cut any better or last any longer.  Just a side rant about coatings:  all coatings REDUCE sharpness.  In certain applications, and filing ain't one of them, their hardness and increased lubricity can allow you longer tool life and greater speeds & feeds, but the hype around coatings make the bicycle industry look like child play.  Cutter performance is all about flute/edge geometry and substrate material properties; coatings can be a performance aid, but choose material, geometry, and sharpness first, then consider the pretty colors.

Where was I?  Oh, I've cut all the hard tubes with files.  I don't typically miter my tubes that way, but I've done it just the same and there's one or two on this least who've even SEEN me do it.  Harder materials will wear files faster, but poor technique is your real enemy.

Don Ferris
Anvil Bikeworks, Inc.
Littleton, Colorado
Ph: 303.471.7533 / 303.919.9073
Fax: 413.556.6825
From: Derek Shuman <dshuman(AT)>
Subject: Re: [Frame] File use

XORider(AT) wrote:
> Once again, please excuse me as I am Dynafile impaired, but what is the
> proper way to file?  Is it about learning how to hold the file parallel to
> the work surface throught the stroke? Could someone please explain hand
> filing or point me to a good book or video on this?  I was never taught how
> to file, so I don't know if I am doing it correctly.
> Dave Schlabowske
Derek Shuman responds:

I will only attempt to add my $.02 to the other excellent postings. I've only made one lugless frame, so perhaps there are better techniques that more experienced people can share. First and foremost, clamp the tube or joint in a tube vise or set of tube blocks fairly close to the miter or
joint. Leave just enough to get your hand around it, unless you are starting the miter, in which case leave as little as possible sticking out. Your ability to rigidly lock the tube or joint off in as many positions as possible is very important, and will pay off in both quality and time saved. Pay attention to your stance; it should always be solid and comfortable, with one foot forward and one foot back, so you can lean into the file. Let your legs and body weight do the work,
not your arms, you need them to maintain rigidity. Look like a prizefighter. Get the height right so you can brace your arms against your body and see what you are doing. You might want one height for roughing, another for finishing. Also, put plenty of light on the subject to see closely what was going on. Try using those clip-on magnifying glasses so you can see close without getting close to the chips or straining the eyes. Don't file when you are agitated or had too much coffee. You gotta be in the right Zen kind of mood for it. For most of the miter, I used two hands on the file, one on the front, and one on the rear. I used a file with a bit smaller radius than desired for most of the miter. Make sure the tang of the file cannot dig into your wrist if it hangs up or your grip slips. I don't like to use handles as they come loose, and you lose some of that feel that tells you how smoothly the file is cutting. If they come loose while you are whaling away, you could really get jabbed badly. There are self-tapping screw on handles that might be good, but I haven't tried them. It might be good to wrap the tangs up with tape, or put a golf ball, piece of rubber tubing or a wine bottle cork on it.  For finishing the miter, I used the proper radius file, found the section where the radius was best matched to the mating tube and went real slow, with short strokes to keep the ideal section doing the cutting. I grasped the file very close to this section with my right,
(pushing) hand and put the right index finger over the section doing the cutting to get even pressure. I would grasp the tube with my left hand, and use the left index finger and/or thumb as a "rubbing guide" to keep the file on track. It might be helpful to wear some kind of glove or finger cot here but I didn't. I didn't have a problem filing away my
finger, but the corrosion problem is something I hadn't thought of. I kept checking the fit as I came down to the mark, and on the top tube, I also checked the alignment between the two miters on the oposite end of the tube using longish sections of mating tubing lightly clamped into
the miters. I would mark sections of the tube with a pencil to show where to shave off a few mils. At this step its tempting to shortcut by trying to file without reclamping the tube in the vise, but avoid it.  Spend the time reclamping the tube and getting a good final cut.  I used this "grasp the tube and guide the file with the index finger" method exclusively when filing down the fillets. Using a rattail file, I
found I could file away with abandon without ever hitting the tubes. The secret is to put lots of light around the joint, preferable small specular lights that give sharp reflections off the joint. This allows you to see exactly where the file is cutting, every single stroke. You should be able watch the fillet grow right out to meet the tube, every stroke. Keep reclamping the joint to put the section your are working on
in the right light. Don't move the lights around, because then you'll have to follow them around, getting into an awkward position. Once your stance is off, you can't concentrate on guiding the file to that exact position needed to finish the fillet without hitting the tube. Have fun, I certainly did.  -Derek Shuman
i miter all my tubes by hand using a
12" or 10" 1/2 round bastard cut file.
From: "Anvil" <ojv(AT)>
To: meade(AT)
Subject: Re: [Frame] Getting the Right Miter Radius
Date: Mon, Mar 26, 2001, 11:18 PM

Approximate coping radii by Nicholson, second cut, half round file size:

8" - approx .5" (1" dia)
10" - approx .5625" (1.125" dia)
12" - approx .625"/.6875" (1.25"/1.375" dia)
14" - approx .75" (1.5 dia.)

The sandpaper around the proper tube is a good trick but doesn't work well
for Tig, it gives an oversize radius.

Don Ferris
From: "e-RICHIE" <richardsachs(AT)>
Date: Wed, 28 Sep 2005 20:46:49 GMT
Subject: RE: [Frame] frame building pictures

i use the 12" to mate the dt as well as the tt to the ht.
i use the 10" to mate the tt to the st.
at the bb, i use the 10" on the front of the st.
at the bb, i use the 12" on the sides of the st and the dt.
-- "Buescher, Conor Joseph" <conor.buescher(AT)> wrote:
Our process is essentially the same, however, use the lug to scribe a
line <cut>

"ours" is too; i scribe the line and then hacksaw & file with
a 10" or 12" half-round as the case may be.
From: "Brian Smith" <briansmithmail(AT)>
Subject: Re: [Frame] Brake bridge mitering
Date: Mon, 24 Jan 2005 19:23:04 -0500

> Hi all, good morning.  What file are yous using to miter the brake
> bridge?  CS's are about 16mm.  Or even methods.
> Thanks, Kurt

I use a round file, about 16"-18" long, and the cut can occasionally end up
slightly smaller radius that the seatstays (that's what you meant, not CS,
right) and if so, 3 or so strokes of a 6" half round remedies that just
Cut to approximate length first though with a hacksaw.

-Brian Smith in N.Y.
From: "Andrew R Stewart" <onetenth(AT)>
Subject: Re: [Frame] Brake bridge mitering
Date: Mon, 24 Jan 2005 19:51:25 -0500

Kurt- Wood blocks and a file work fine. I think the 16mm stays are about the
same as my 12" rat tail (round) file. Andy.
From: "" <brianbaylis(AT)>
Date: Thu, 12 Aug 2004 05:10:55 GMT
Subject: Re: [Frame] (flying off the) handle


If you come to my shop you will see that many of my files have just a
"bulb" of masking tape on the end. I find this more comfortable for me
than some of the larger file handles. I'm kind of on the short side and my
hands are somewhat small, so this works for me. It may not work for you.
Cheap, fast, easy. Also an excellent addition to a needle file.

Brian Baylis
Date: Wed, 22 Jun 2005 09:27:44 -0600
Subject: Re: [Frame] How tight the tolerance for a miter?
From: Harold Bielstein <hkbielstein(AT)>

Here's what works for me to get a nice tight mitre. Chuck the mitered
tube into your vise in your tubing blocks so that the mitered valleys
are in a horizontal plane. Wrap a short length of tubing that properly
fits the miter with 180 or 220 production cloth and use it as you would
use a file back and forth across the mitre. Do this a few times till
you cant see light between the mitered tube and its adjoining tube.
From: <davebohm(AT)>
Subject: Re: Re: [Frame]  TubeMiter exe and other ramblings
Date: Thu, 26 Aug 2004 17:03:47 -0400

I don't really use any drawing at all.  I know I am not the only one.  I use some sort of spreadsheet that I believe came from Mark Bulgier or something and it just gives the pure numbers.  tube angles and lengths and I translate lengths with a 4 foot dial caliper  (thanks don ferris)  works

Dave Bohm
From: "Neil" <info(AT)>
Date: Sun, 11 Jan 2004 00:57:43 -0000
Subject: [Frame] top tubes - one way to get it right.

Hang in there Jason, getting the top tube right is one of the hardest parts of the game if you're working on your own. I had the advantage of working with some of the best frame builders Britain ever produced, and watched intently as they went about their stuff. Now and then, they'd take the time to show me how. This was one method, which is a good one if you're starting out. It uses pins, which have been mentioned elsewhere; I just new them as brazing pins (coz that's what they were called, and Holdsworthy supplied them by the pound) but I'd say woodworkers 16g panel pins would do the job. (although they're probably called something else in the USA!)

First, slide the head lug over the tube and scribe around the inside. Take the lug off. Do a rough miter to the scribe line. Slide the lug back on and check your miter. Slide in the head tube and check for fit. Check against your measuring board or in your jig for angle. Heat and bend the tangs to match the angle if needed. If you're happy with this part, drill a pin hole in the lower tang and lightly tap in a pin. Slide the seat lug over the tube. Measure the length, centre to centre. When you hit the desired length, check for seat tube/head tube alignment. When it's as near as you can get, drill and lightly pin the seat lug through the lower TT tang. Scribe the tube around the inside of the seat lug. That's you got the right length seat tube. Now you can remove the pins and lugs, cut and file your miter, deburr the inside of the tube, etc. Now, with the pins out and the lugs off, check everything is tight against the head tubes and seat tubes. Make the most effort to get the head tube end right, the seat tube end isn't as crucial. Fine tune the miters. Heat and bend the tangs for a tight fit if you need to. When you're satisfied everything's as close as can be, fit the lug to the tubes; bang in the pins to fill up the holes; do your fluxing and get brazing. Tack the head lug at the front of the head tube; tack the seat lug at the back of the seat tube above the lug line if you're using a torch. If you're using gas/air and a hearth, pin the seat tube/seat lug.

There are other ways to do it, but this takes a lot of beating if you don't have a data book full of mitered tube lengths.

Good luck,

Neil, London, UK
Subject: Re: [Frame] Tube Coping
From: meade(AT)
Date: Wed, 30 Jul 2003 15:21:02 -0400

>What is the easiest,
>cheapest, and less wastefull method?

I second the idea of files.  It's not hard to do, and doesn't
take much time, plus you don't run the risk of chewing up
a tube if something goes wrong in your lathe, and you'll be proud of your handiwork when you get a nice fit.   I use the method outlined in Paterek's manual:  If you use lugs, use the inside of the lug to scribe a rough outline (I use DyKem to make the scribe easier to see).  Then use a hacksaw or a rough file to chew out the bulk of the material, and halfround files and small files to finish the job.  Use a machinist's square to make sure the points are the same height (helps to ensure that the miter is centered on the tube), and use the mating tube and a bevel protractor to make sure the angle is correct and that there are not big gaps.  You can even wrap sandpaper around the mating tube to make really small changes.  So far (four frames) I've never messed up a miter this way.  It takes me probably 30-45 minutes to get a really nice miter with not much daylight showing. If you decide to go lugged, there are only three miters that require any precision at all.

Alex Meade
Lexington, KY

From: "Al Andersen" <andersens(AT)>
Subject: Re: [Frame] top tube woes
Date: Tue, 25 Feb 2003 17:45:57 -0700

If you are making a full size drawing, lay out the tt in it's
posistion, and measure the angle with a protractor.  You could do trig calculations too.  Then take a T-Bevel and lay it on the tt st angle and set it to that angle.  As you file the tt check your progress with the T- bevel.  I have done my frames following the Paterick Manual.  First miter the tt to the ht.  Have the bb-st and ht-dt done, together and straight. Then fit the tt-st miter till it lines up according to your drawing.(Without messing with the tt-ht miter you did first).  Expect this to take some time.  When the tt drops into place, braze/weld/glue, whatever, and you got it!  Does this help?  Kurt andersen
From: meade(AT)
Subject: [Frame] Re: [Frame]file size
Date: Mon, 9 Dec 2002 18:13:00 -0500

>Hi all, what size file aproximates a 14mm dia. ss tube, as in mitering
>the brake bridge? What do you use?  Thanks Kurt

I use too-small round files to rough it out, followed by sandpaper wrapped around a piece of seatstay.   Makes a slightly-too-big miter, but it seems to be fine for silver brazing, especially if I'm using the reinforcing tabs beween the bridge and the stay.  If I'm feeling really picky and not using the tabs, I wrap the sandpaper around a smaller diameter part of the stay (at this point of course the stay is part of the frame !) to get a "perfect" fit.


Date: Wed, 25 May 2005 09:58:25 -0400
From: Doug Fattic <fatticbicycles(AT)>
Subject: [Frame] How to cut out lugs

Some on the list might have wanted to cut out their own lugs at some point in time so I thought I would post some detail on how to do this.  It involves two basic steps.  1. Getting the design on the lug and 2. Cutting that design into the lug.

I start by unrolling the shape of the round tubed lug, flat.  There would be two parts to each lug.  I cheat by taking a piece of tubing that is the same size as the lug, miter it so that it has a sharp edge and wrap a piece of paper around that miter.  Now I cut the paper hanging over with an exacto knife.  Most of you aren't going to have these components available so just wrap some paper around a lug, trace the joint with a pencil and cut it out.  A little back and forth and you have the shape needed.  A little help here, does the mitering program some of you use create this shape?  That would make it simple.  Once the shape is created - the one end being the shape of joining the other tube and the other end being square - the design can be placed on it.  Of course those that know CAD can do this on their computer. However for the more computer challenged, this can be drawn by hand.  One trick I learned is to blow up this size 4 times.  Our school copy machine can do this.  This allows for easier drawing using French curves or whatever.  Once your Picasso quality creative drawing is finished, the copy machine can reduce it back down to normal size.  Now I use sticky back clear film in the copy machine from the drafting supply store.  This stuff has two parts with a backing paper covering up sticky.  The Xerox machine will print your design right on this paper.  The backing paper is pulled off and the clear film is applied to the 1/2 section of the lug being worked on.

Don Ferris's (Anvil) lug holder is a real bonus here.  He has made a very nice one.  I forget the cost but I think it is $275.  The other tools you need are a jeweler's saw and blades.  Exacto makes one.  These are available at jeweler's supply stores.  Look under this heading in the yellow pages. South Bend, Indiana used to have one and so does Toledo, Ohio.  Does anyone know a mail order supply house for these?  You also need to get a bunch of blades.  You can go through these like water.  They cost about 35 cents apiece and the brand you want is Herkules.  They are sold in packs of a dozen or gross.  They come in all different sized teeth.  I think I commonly use 3/0 size.  Sometimes I want bigger blades like 0 or 2 and sometimes smaller like 5/0.  I keep a bunch of different sizes around.

Now you need Swiss Pattern Jewelers files.  They often times come in a set of 12 shapes but you don't need or use all shapes.  I would get the 0 cut, 5 1/4".  The ones I use are the Half round, half round with a safe edge on the flat side (handy in some tight places where you don't want to disturb the other edge), round, knife and I forget the name but the file edge (teeth) are on the flat side and the back side is safe (without teeth) and triangle in shape.  This allows you to get into tight places.  I seldom if ever use other shapes except maybe the one that is flat with out teeth and the edges are rounded with teeth.  This also works for the bottom of the seat tube binder cut.  Although the 5 1/4" do the bulk of the work, I also have some 4" ones for some situations.

Now when you cut out the pattern, you don't cut right on the edge.  Your blade is coming in at some awkward angle so you leave a little space.  That angle also makes them break easily.  Oh first you might drill out some places that the drill size matches the design.  Before going too far back with you cutting, you start filing to the edge of your design lines.  This is because you want the supporting strength of a solid lug behind what you are filing.  So you don't cut out the whole thing and then start filing to the edges but work back and forth cutting and filing.  At some point, you need to emery off the design to see if it looks right and so a little more touch up filing to make it right.  Next you are going to admire your work and think to yourself, Doug was very generous in telling me the secrets of how to cut out lugs.  I'm going to be generous back and write him a check right now for $100.  No need to send it Priority mail, I'll be happy anytime I get it.  Or more realistically just send comments and thoughts.  I didn't take a lot of time writing this as I have a bunch of bikes to paint today so some detail can be expanded on.

I should mention that I have some blank lugs.  I'm not giving them up easily (or maybe not at all - they came from England in the 50') but if someone makes a great design, I'll negotiate.

Doug Fattic
Niles, Michigan
Date: Sat, 10 Jan 2004 17:38:41 GMT
From: brianbaylis(AT)
Subject: [Frame] The two basic assembly methods

Fellow framebuilders,

Sorry it takes me so long to get to this stuff. I really have a lot to do. The purpose of this information is to aid the new framebuilder in understanding what basic methods are in use that relate to the small time builder. Jay asked what methods we all use. As far as I know, there are these two common practices that work well for our needs. The pinning method and the tacking method. I personally use the tacking method with a few minor variations where I use a pin for "cheap insurance" against movement during free brazing.

My experience is that most veteran framebuilders seem to use the method they were exposed to when they were taught or learning the craft. Back in the old days, many of us learned from either English or Italian framebuilders in one way or another. I learned the basics of framebuilding from Masi during my employment at the Masi USA Carlsbad shop. There the tacking method was used.  I have modified the sequence considerably for my own needs and taste, but basically I tack the frame as opposed to pin it. I don't know that much about the pinning method and I haven't been to Richards' site, but I think I understand the jist of it. Every framebuilder will probably end up with a system that has been taylored to their fixtureing and other needs based on the circumstances. Here we go!

To begin with, both of these methods require that the frame do into the fixture with absolutely NO stress at any of the joints. That means that all of the angles of your fittings must agree with the dimensions you are seeking to accomplish. The proper angle at the BB shell is critical. Make sure you can assemble the frame completely and it will hold itself together at exactly the angles and dimensions you require before you attempt to put it into the fixture. I work off of a full scale drawing which allows me to check and double check everything as I miter tubes and prepare fittings for joining. I also then set up the frame in my fixture without any lugs to check miters and the fit of everything. Then I will set it up again with the lugs and all to make sure everything is perfect and in plane and I'm satisfyed  that all systems are go. I then disassemble everything and clean and flux everything before going back for the final setup and tacking operation.

>From what I understand of the pinning method, the frame goes through the same basic preparations I just described and then assembled, fluxed and ready, in the fixture. Then holes are drilled in key locations and the pins wedged in. The frame is pinned like a porcupine until it is held together completely. It is then removed from the fixture and free brazed.  Then the frame is aligned and that is the end of the process.

The tacking method goes thusly. The tubes and lugs are fluxed and assembled in the fixture. Then the lugs and tubes are tack brazed using the same material you will finish braze with, in generally three places on each tube. The frame, which is now a solid unit is being held together and in dimension and alignment by the tack brazes. The frame is then put on the alignment table and checked for perfect in plane. If any movement is required, generally only light finger pressure is required to move it back into place. It came from where you want it to be and it will go back there if your fixture is doing its' job. Once perfect, the frame is free brazed. After brazing the alignment is checked again. Nothing should change beyond a few thousanths if you use low temp. silver braze material.

A certain amount of how to do things depends on what sequence and methods you use to assemble your frames. Some frames come out of the fixture whole, with the rear triangle already attached. Masis came out that way as do Richards' frames, from what I understand. Eisentraut used to (and still may) build a front assembly of the TT, HT, and DT in one fixture and used a second fixture to attach that to the rear of the bike. Mario Confente would start with a DT and HT held in a simple fixture at the perscribed angle and then move on from there piece by piece sort of.

I like to build just a main triangle in my primary fixture. I make a sub-assembly of a completely brazed and aligned seat tube before I put the other tubes in the fixture. I go through the same sequence. Put the BB shell and ST in the fixture and tack it. Check alignment, braze, recheck. That gives me a foundation of the centerline of the frame that is stable and perfectly aligned to set up in my fixture. Also it helps guarentee a perfect fit and braze joint at the most critical junction of the frame, the DT and ST at the BB shell. I see a fair amount of frames where the miters and brazing are sub-par here. This allows me to control this area completely, see the complete penitration at the seat tube which in turn will almost guarentee perfect jointery when the DT is brazed because the silver already there attracts the new material. I do this as "cheap insurance" to a perfect joint in this important location. Once I have a perfectly aligned and solidly brazed main triangle I have another fixture that I use to attach the main to the rear triangle. The  most important time for your frame to be in perfect alignment is while it's in the fixture. Attend to that and the rest of alignment is a snap. There are a few reasons why I have gone this way even though I learned from Masi who puts the entire frame in the fixture with the two stays already done as a sub-assmebly. One, the way I file my lugs requires that I finish file the seat lug before attaching the stays. That's how I get the look around the top of the seat lug that I do. Most bikes are thick where the seat goes in, I prefer a different look. Also, because I monkey around with all sorts of different variations in seat stay attachments, it allows me to experiment and be artful if I want to do that. It also allows me to have more control over the alignment of the rear dropouts. Working with the whole frame at once seems like too much to keep an eye on.

So, there you go. There are as many was to build an excellent frame as there are excellent framebuilders. The most important thing is how careful you are as the operator. Every method has it's quirks. Each guy irons out his own bugs over the years. It's an ongoing process, up to a point. Some methods work better for small timers and hobbiests and some things are better suited to more production.

Brian Baylis

1 comment:

Anonymous said...

Very best site. Keep working. Will return in the near future.