Cassius

I've just recently really gotten into cycling. It started with the fact that I've been commuting by bike for a couple of years but never had a bike the correct size or one that was in OK condition. I started to look at the commuting as a godd excercise (lika many others I imagine) so I bought a used steel frame (A Columbus Aelle frame) and some parts, put it all together and wow what a difference..

I've always liked to do things myself, hence why I bought a frame and parts instead of a complete bike. I've also always liked the older more classic stuff (I'm mainly a car guy and mostly interested in cars that have motorsport heritage from the 70īs and 80īs). So from getting into cycling to thinking about building my own lugged steel frame wasn't a very far step to take. I'm comfortable with most tools found in workshop's but haven't mastered the TIG yet so I thought a lugged frame would be a good start.

Now.. I'm 1.93m tall and weigh rougly 90kg with an athletic build, so my main questions are if the tubing I've looked at could be considered strong enough for an amateur or if I should look at something with thicker walls. And if the type of tubing is usuable for building a lugged frame.

The tubing I've looked at is Columbus Life (Looked at Tange and Vary-Wall aswell, reynolds doesn't seem to be that easy to find). I've read that Reynolds 753 which is heat treated isn't something a novice should start with so the main question is if Life tubing is better suited? And if not, then why?

For tubing numbers I had planned to use:
TT 28,6mm 0,7/0,45/0,7 wall
ST 28,6mm 0,7/0,4/0,6 wall
DT 31,8mm 0,65/0,45/0,65 wall (there is also a version that is 0,75/0,45/0,75)
HT 36mm 1,1 wall, or if I really want to chase weight a Vari Wall 35,8 with 0,9 wall.
Chanstays 30x17mm oval 0,7/0,5 wall
Seatstays 17/12,5mm 0,5 wall

The angles as planned at the moment are (Long shen lugs):
73 degree seatlug
74 degree top lug


58 degree downlug

59/62 degree BB shell
Drop out should be 72-74 degrees something.


BB drop 78
Wheels base right aroung 996mm
Fork 365 ish with 40-45 rake (should give 57-52 trail)

The "planned" tubing and lugs, with weights adjusted for the lengths of the tubes, end up to 1616 grams, So in my world this should end up being a sub 1700g frame (havenīt acounted for mittering the tubes, on adjusted to the c-c lengths). That about 650-700g lighter than the frame I have I belive



Size 58cm square (I have short legs and a long back for my height)

mikeread

Hi Cassius

I am 3 inches shorter than you and 16kg lighter.

My fist frame used similar size tubing to your planned build and slightly thicker gauge (8/5/8 ) The exception being that I used a smaller head tube for a 1" steerer. Reynolds 631 tubes with 725 stays.

I have been riding this bike for 15 years and have been very pleased with it, it is very comfy to ride and is excellent for long distances or commuting. However, load it up or ride it hard and it does start to feel a little flexy.

I recently built my second frame from Columbus Life double oversize tubing 8/5/8 DT 7/5/7TT 8/6ST 1.1HT and whilst being a little harsher, is much better for riding hard.

The recent frame weighed 1992g the first frame 1996g - go figure that one out! This is probably because the Columbus life stays are much lighter than the Reynolds ones.

With your added size and weight you might find your proposed build a little on the light side - but I am no expert, wait and see what they suggest :-)

Good luck with your project and enjoy it.

Which side of the pond are you? Just notice you are using metric.

Cassius

On the "right" side (Sweden)

The flexiness is what I'm concerned about.. Yeasterday I took a new personal best to work (It's straight throu central Gothenburg, a total of about 35 crossings), my average pace was 31.2 kmh with a top speed on the flat of 48kmh (thats rougly an average of 19,4 mph and a top of 29,8mph). This is over 10,5km or 6,5 miles.

During just the first month with this bike I've cut the time to work with 4+ minutes! (The "old" bike was a medium sized Nishiski Hybrid with suspended fork weighing 16+ kg...).

So I wouldn't say I'm a slow rider, but maybe not a fast one either

What tubes are more important to beef up concerning flex? I don't want to go up more in tube sizing, keeping these sizes I should get a frame that looks like a 1" standard frame in size 52, but in size 58 Wall thickness is less important. I just want a bike that I can evolve on for the coming years, that's still lighter than others store-bought-full-carbon bikes costing 3-5 times more (If weight was my ONLY concern, I would probably buy a chinese carbon frame. But it's not )

unterhausen

most people recommend more substantial tubes for a first build.

Nessism

Not only are more substantial tubes easier to build with they will make for a stiffer frame; stiffness is proportional to wall thickness. Columbus Zona or Deda ZeroUno are both excellent choices for lugged builds. All up weight after paint on a 58x58 frame should be about 1900 grams.

Andrew R Stewart 

Don't make the mistake of connecting frame flex or weight with creating a faster (as in actual time to distance, not moment by moment perception) bike. There's sooo much more to how a bike/rider goes faster (or slower) then weight or flex.

I completely agree with Eric's suggestion to use thicker walls for the first (few) frames. Besides the aspect of overheating during the joining (especially with a lugged construction) the tubes have far less material that can be sacrificed under a file or sanding/blaster. Unless one has pretty good fabrication skills already most mentors will advise .8 or .9 tube walls at the joints for the first frames. Andy

TiHabanero

How about a rookies input here? Building a forth frame at the moment, so certainly meet criteria for rookie. First frame was Reynolds 531, good old 9 6 9 walls. 2nd frame was OX Platinum, I believe the walls were 7 4 7. Lugged construction, and it went together and rides fine. One word of caution as Andy pointed out. be very careful with files and stones. not much there to begin with and a deep nick can compromise the tube wall.

Have nicks in several spots on the Reynolds frame and after almost 30 thousand miles, it is doing just fine. Don't overthink it, just do it, albeit with an bit of caution when taking tools to the thing.

Cassius

Well to be honest... compared to building my Opel Kadett (also known as a GM "T" body, or Chevrolet Chevette/Pontiac Acadium.) with a different rear suspension and axle, altered front suspension, new steering rack, dry sumped V8 engine with ITB's and 180 degree headers (I've even painted this myself).

Building a lugged bike frame feels like a piece of cake

I agree that thinner tubing will be harder to heat to the "right" temperature since it will heat up faster. Depending on the material thickness of the lugs that could be a bit tricky. But whatīs life without a challenge every now and then? I'm fully aware that I will have to give the brazing time in order to not overheat the tubes and lugs, and from what I've gathered so far heat treated tubes like Tange Ultimate will require even more skill and possibly even an after treatment? While air hardening steels would be easier to work with. On the car I've used both 1mm and 0,7mm sheet metal for the bodywork and when welding them you definatly need to be more careful with .7 to not burn a big hole in it.

Larger diameter tubing adds stiffness, so a 25.4mm 7-4-7 tube will be weaker than a 28.6mm 7-4-7. Hence why I aimed for 28-28-31 with a 36 head tube instead of 25-28-28 with a 31 headtube. I haven't ridden any bikes with super thin tubing so I can't tell how this "flexiness" might feel which is why I'm asking for tips and recomendations

Comparing "std 1 inch" tubing diameter to OS that I'm thinking of is done here for example: thetallcyclist.com/2016/09/theoretical-framebuilding-part-2-tube-size-and-butting/
There it looks like an OS frame is roughly twice as strong as a std frame all things equal (or am I misreading the info)? There is also a mention of tube butting compared to strength where a straight gauge 0,7 tube deflects about 50% less than a 7-4-7 and a litle less than half of a 5-3-5 tube.

Rough estimate should give that a 0.7 straight gauge 1" std frameset is as strong as a 7-4-7 tubed OS frame. This still doesn't tell me anything about how much flex is OK and not.. But it does tell me that "my choice" in tubing should be as strong as a 1 inch frame with straight gauge 0,7 tubing (roughly). The Aelle frame is straight 0,8 and I can't detect any flex. Is it wrong to assume that a similar frame with 0,7 straight would flex just a tad more?

And if I decide to stick to these thin walled tubes but wanted to upgrade the tube that does the most to improve frame stiffness. Which tube should I then choose? I'll take another look at tubing alternatives when I get a chance But for now I'm just soaking in as much experience as I can from others

squirtdad

OT but have friends in göteborg. Most went to Chalmers and one commutes in Velocar (think he works at Abscom). Nice town

Doug Fattic 

The reason I don't recommend my frame building class students use .7mm tubing is not only because of the problem of heat control keeping the the thin tubing within the necessary temperature window but also because it is very likely to be heat treated and one has to know how to heat the joint evenly so the frame doesn't get out of alignment. Bicycle tubing will move a lot when heated and thin tubing moves more. Thin heat treated tubing does not bend much (if at all before collapsing) so it is difficult to impossible to bring the frame back into alignment once it is out of whack. Non-heat treated thicker tubing has a lot more margin for error and correction.

Having given you the proper warning, I like the ride of thin wall tubing a lot more. It has a lively feel that I don't get with the standard stuff. My warning should also include the observation that only the top 5% or maybe 10% of students are anywhere near good enough to braze thin walled heat treated tubing. An experienced brazer has a lot of their motions in muscle memory so they don't have to think too much and can concentrate on even heating. Of course I don't care if you end up with a crooked frame with cooked joints.

unterhausen

one thing nobody has mentioned is practicing brazing. You can get aircraft tubing that will nest on the next size down. Pretty sure that English diameter tubing is available in most of the world. So get some 1" tubing and then some 1 1/8" .058" wall tubing and practice brazing lugs.

I think you should plan on throwing the first frame away and not try to build your dream frame right away. But that sort of thing is up to you. If you mess something up, individual tubes are available. Getting non-heat treated tubing is a good idea, and I'm pretty sure the thinnest tubing you can get in plain 4130 is .8.5.8, which is plenty thin

Andrew R Stewart 

The weight savings between two frames with a .1mm less wall is pretty minimal. Add a OS step up with that .1mm less wall and the weight is nearly the same. But still way less then other elements that make a bike complete. Of course an OS step up brings a stiffer tube. IME it's more the fork's steerer that speaks to much of the stiffness of real riding. Not just out of saddle sprinting or hill climbing but cornering and how the bike deals with non rider weight (like bottles/bags and their placement). I've mentioned the importance of forks before, the difference between a 1" and 1 1/8" steerer is noticeable. Then there's the wheels contributions...

So many feel that complete fill (with lugs) is the goal, yet millions of production frame challenge this assumption. The comments Doug mentions are of greater value IMO though. The heat cycle, time to temp/, the duration/and evenness about the joint are critical for alignment and not loosing the initial tube's strength more then needed. It is this heat cycle that is hard for many to get a handle on. A common problem with newbies is to use a small flame (to better control overheating) but ending up with a very long at temp duration and more unevenness of heat about the joint. If a large flame is used then the overheating thinner sections becomes more likely. So with any highly skilled endeavor the practice and review are strongly suggested. Eric makes a real important point about the value of practice. Andy

Cassius

Doug: Heat in this case will be every beginners problem and I appreciate the advice! This is why I'm asking hese questions
The Columbus Life tubing however I can't find any information on whether it's heat treated or not? Nor the other tubes in the Columbus range. Most of the Columbus range come in .7 butt wall thickness, there are some .8 and only a few .9 from what I've seen.

I took a look at Zona tubing for the main triangle with 8-6-8 for top and down and single butted 8-6 for the seat tube. With the same lugs and stays that would be 142 grams heavier. Adding chain and seat stays in Zona the number increases to 201 grams (with the same tube length compensation as earlier) That ends up as 1817g without taking the mittering of the tubes or any other braze onīs into account.

But how big is the difference between .8 and .7 when it comes to heating? When I've welded sheet metal on the car and compared 1mm with 0,7mm I can use the same current for the weld, but maybe only hold it for 3/4 of the time before I burn a hole. These spot welds when I get them right are almost completly flat.
However, that tells me that the difference in heat management between .7 and .8 shouldn't be that big. Given that the material is exactly the same of course. I'd say that I'm more afraid of underheating the connection than overheating to be honest :/

The temperature needed is also connected to which braizing material you use I guess?

Unterhausen: Fair point. I could buy a piece of the cheapest tubing in the same wall thickness as I'll use for the frame where I get all the other tubes and do a couple of practice runs before going to the frame!

Andrew: Then I guess you would agree that going up to OS tubing and steerer is a smart choice?

If I understood you correctly, would you say this method would be a "smart" begginners method: Take out holes in the lugs so that the distance that I need to "pull" the brazing over becomes shorter, and do only parts of the lug at a time. For instance do the top lug to the top tube first, let it cool off. Then the head side of the lug? It gives more heat cycles, but shorter ones. From what I've gathered so far, the heating should be kept to a small part of tube right at the lug and not spread out over the tube to get the heating more evenly increasing over the tube lenght? I guess this is to keep molecular movementas concentrated as possible so that the steel wont shrink/pull over a longer section of the tube?
Wouldn't pre heating the entire tubes in say an oven make this process easier in that case? The heat difference between the non heated parts of the tube will be smaller (bronze melts at almost 1000 degrees C, preheating the tubes to say 200C is 25% less heat difference) giving a larger window for error?

By the way, silver or bronze is mainly used for brazing. Is there a reason for this? Aluminium for example melts at a lot lower temperature. Is the 950-1000 degree celsius needed for something else than actually getting the "filler" to melts throu?

Doug Fattic 

The problem with using 7/4/7 is because it is almost always heat-treated and therefore can’t very well be bent back into alignment. So the difficulty of using it is based more on the probability of using an uneven heating pattern than overheating it. On the other hand 8/5/8 top and down tubes are available in non-heat treated versions so it is easier for after brazing alignment adjustments. An option for you to consider is using an 8/6 seat tube (there is not much difference in weight between 8/6 and 9/6 seat tubes) and an 8/5/8 non heat treated down tube and then if you still insist, a 7/4/7 top tube.

Let me tell you a story involving the history of modern American builders and the problems of alignment. This will take a few minutes so I ask for your patience. Those of us in the States that wanted to learn how to build frames in the 70’s had to learn from Europeans (with Andy Stewart’s teacher Albert Eisentraut being the exception) or figure it out on their own. The ones that knew how to do it before WWII had died or retired when the bike boom started in 1970 and there was a fresh demand for custom frames in the US. Around 1977 Reynolds came out with its thin wall 753 tubing that was heat-treated and therefore not very alignment adjustable. Thin tubing is less likely to dent if it is heat-treated. They didn’t want their new flagship tubing getting a bad reputation from hack builders not knowing how to properly build with it. For example it was common in the classic era for builders in the UK to have very modest equipment and use brass (Americans say brass even though it is really bronze). At the time heat-treated tubing needed to be brazed with lower melting temperature silver so it wouldn’t anneal the temper of the tubing and take away the advantages of heat treatment. So what Reynolds did was require frame builders to pass a test if they wanted to buy it. In the beginning this test involved building a whole frame that was tested for alignment and then cut apart to see if silver had penetrated everywhere. In 1978 when I was visiting the factory in Birmingham they told me that every single American builder that had taken the test had failed. Every single one. I don’t know which of my contemporary colleagues had submitted a frame for testing and failed (I hadn’t tried yet in case you were wondering) but I get the idea it was quit a few. It should also be understood that they had already been building frames for several years and they had most likely been taught by somebody that knew more than nothing. It is with this background that we are naturally skeptical that on your very first try you are going to do a good job using thin wall heat-treated tubing.

There are ways that a rookie can build a decent frame on their 1st try but that involves quite a bit of instruction and practice. It is the reason I've been teaching frame building classes for 40+ years.

Andrew R Stewart 

I would never suggest brazing a single socket (lug hole) in more then one heat cycle for a few reasons (with exceptions I'll mention later on). The heat cycle starts when the flame touches the steel and ends when the steel is back to room temps. To do this multiple times on the same joint is to add up the collective time the steel is at temp, remember that to flow the filler and have it wet out to all the surfaces it is "suppose to" means that the area of steel at brazing temp is greater then where the filler is being added to. For the filler to "mingle" with the previously done then cooled off filler that first filler needs to also be up to temp.

Most everyone who does this stuff (and teaches it, I've actually had 4 people who I regard as my mentors/teachers although Doug is being modest as he's one of them) will instruct the joint to be brazed in one cycle. Ideally this is done when the entire joint is up to complete brazing temp, so that the filler can flow through all the surfaces in one session.

Back in the day and before compressed combustable gas was readily available one technique was to use the hearth method. The joint was prepped and placed with a wall of fire brick surrounding much of the joint. A large jet/river//flow of flame (gas light gas when street lights were gas) nwas aimed at the joint and the firewall retained the hot flow all about the joint at the same time. Filler was applied and would flow through the joint in one step. As compressed combustible (and oxygen) gas became available a hotter flame was also available and the flame could be directed into a smaller spot. The convenience of this method won out over the "broad paint brush" method which was far less portable as well. here in the US the use of acetylene/oxygen became The WAY. Certainly the other industries that use brazing did the same for the most part. If more complete and even heating was vital then multiple flames (of an oven) was employed. The small framebuilding shops learned that using a larger AC tip, having a larger (not hotter but with greater total energy) flame allowed more of the joint to be brought up to temp quicker and with skill the filler's flow was complete faster and more evenly. Less "walking about the joint" with the flame's inner cone was the result. The joint had more even heating over a shorter time. One of the challenges for a newbie is to achieve this goal, the more even heating of the joint for less time. But a newbie isn't as skilled and usually wants a smaller flame so they can keep up with what's happening. So newbie's joints are often unevenly heated (think distortion) and at temp for far longer. During the brazing the different areas of the joint will raise and fall in temp as the flame is directed at relatively small points. It takes a lot of practice for most to get past this slow and cool method and become skilled at the better method of a big flame and working faster. One reason that propane is liked more and more by the new builders is that it's larger and "softer" flame reduces the pin point heating nature that an OA flame tends to have.

Now back to exceptions.

One is when fillet brazing. many will do a joint in two steps. The first is a flowing of the filler into the tubes' contact surfaces. This interface is very small, just the mitered end of one tube against the surface of the second tube. With only a mm or two of interface the flowing goes very quickly (compared to flowing a lug). The heat affected zone remains close to the joint as the time for the heat to travel away from the joint is short. The at a later time (which could be only seconds or more then a day) the fillet would be built up on top of the flowing brazing. The fillet is typically done at a slightly lower temp so only the surface of the tube/flowing braze will be up to wetting out temp. Yes the joint is not as evenly heated like when flowing a lug but a skilled builder will add the fillet in sections about the joint to counter the distortions, just like when welding BTW. Seeing a well done fillet is a thing of beauty.

Another exception is when working on a BB shell joint(s). It is fairly common to first joint the seat tube in one session, Then later do the same for the down tube and later still add the chain stays in a third brazing session, But the shell is fairly thick walled so strength losses are not considered an issue and the tubes adjacent to what's being brazed don't typically see the temp levels that are a concern. Still some builders will do the ST and DT in one session, or even all the shell's brazing at the same time.

The last common exception is when adding the seat stays to the ST/TT joint. This step is often done after the ST/TT joint is fully brazed and also cleaned up. It's vastly easier to both clean up the ST/TT joint without the stays getting in the way as well as to control the rear end's alignment better.

As to the joint sequence some will set up the entire main triangle and braze it in one session and others will do a joint at a time and between correct their alignments one joint at a time. Doug has taught this joint by joint method to his students often. (Trying to be careful as to put words in his mouth) this slows down the work and breaks up the amount of "stuff" going on at any one moment.

My own method is usually to join the ST to the shell then prep the rest of the main triangle and tack the joints. Then check for alignment and complete the brazing of the rest of the joints with the frame out of a jig. I will braze one joint completely at a time, working my way around the triangle with the ST/TT for the end. Sorry if I didn't make myself clearer earlier. Andy

unterhausen

silver and bronze have roughly the same strength. Aluminum wouldn't wet out and you can't get it to flow. The trick with heat treated tubing and silver is to keep the temperature low. The metal should be roughly pinkish. This keeps the heat affected zone small. Also, since you might need to align the rear triangle, do everything but the seat stays first. This allows the rest of the frame to be moved around. The seat stays couple the chain stays to the front triangle, so you can't move either end of the frame without moving the other end.

Nessism

It's much easier to learn how to control heat than it is to learn how to build a straight frame. Particularly if you don't have a fancy professional grade frame fixture to build with. So that means some measure of frame alignment coldsetting will be needed, and with thin tubes there is more risk of buckling, particularly if any torsion is involved such as is often the case where the head tube is out of plane.

Repeating myself but .8/.5 Zona or Zero Uno would be about perfect for a 58x58 frame. It's reasonably light, reasonably stiff, and reasonably easy to build with. The .8 butts are reasonably short which saves some weight too. You can go to Life chain stays and seat stays (although at 17mm the seat stays are a little fat) if you want to save some weight. Plug-in type dropouts are light and easy to build with so they are a good choice. Building around a carbon fork is another good option if you are paying attention to weight.

Andrew R Stewart 

Just so there's no assumptions or miss understandings about jigs and alignment. The jig is not responsible for the frame's being in plane, the two wheels track the same and are centered in the frame/fork. At best the jig does a better job at reducing the errors that sloppy prep can create. It's generally considered that proper mitering and heat control does the job of maintaining tracking. Andy

TiHabanero

"It's generally considered that proper mitering and heat control does the job of maintaining tracking."

Learned that the hard way on the last frame I built! Tack and align, tack and align!

Nessism

With a good frame fixture it's MUCH easier to assure good miters, with everything in plane. It won't assure a straight frame in of itself, but it will make the job much easier.

duanedr 

I think the crux of your question is whether the proposed tubing is too flexible or not. Aside from the issues of heat and joining thin walled tubing and annealing HT tubing etc. I can add this:

I was relatively big racer - when I was racing, I was about 85kg and roughly 1.78m tall Cat 2 sprinter type. My frame tubes would be somewhat shorter than yours as I'm a bit shorter. I found that after I quit racing, using a 35mm .6-.4-.6mm DT and 28.6 .7-.4-.7 TT and 28.6mm .8-.6 ST with 36mm HT felt really good (25.4 TT was too flexible). The thin walls weren't good for dent resistance (the frames were test mules), but the snappiness felt great! So, taking that as an example...

From the models I have that compare the torsional flex of a tube of various diameters and wall thicknesses, the 35 x 6-4-6 DT has a torsional rotation of 3.09* with 100NM of force applied. What you are proposing has rotational deflection of 4.22* in the same length as I have so, maybe extend it 2cm and the rotation deflection jumps to 4.54* or nearly 50% more rotational deflection than my preferred 35 x 6-4-6.

Obviously, there are a lot of variables here and I am not saying that the tubes you propose will result in 50% more flexible frame than what the rest of the group is proposing but, I believe this is a reasonable way to assess the relative impact of tubing choices on frame feel. Forks, butting profiles, wheels, Geometry, tires, chainstays, tire pressure etc all play a part in how a bike feels.

What you propose will be much too flexible for the rider the bike is being designed for (and as discussed, the brazing will be less forgiving of a new builder).

If your downtube has to 31.8mm, a tube with .9-.6-.9 walls has roughly a 3.10* rotational deflection in the longer length which is nearly identical to the 35mm 646 walled tube I have found to be very good feeling. the 9-6-9 tube is heavier but - 200gm?! The dent resistance will be better and it will deal with heat better than what you propose.

Cassius

First of all, thank you all for very educating and understandable answears! Nine times out of ten the answears you get on public forums are more in the line of: "No that's a bad idea", "only pro's can do that", No don't do it that way/use those parts, those are crap! Do it this way/use these parts instead" without any explanation as to why they have the opinion that they have. The tenths time you get the explanation. This however must be the eleventh time (och maybe one in a hundred) with both the explanation And tips on how to evalute how to tackle "the problems"!

Doug: So wall thickness isn't the real issue, it's the tempered steel that's the problem? (Due to the odds that a rookie won't be able to control the temp good enough)

Andrew: Great tips on the brazing process. The mitering will be a challenge, since I don't have a good jig for the hole saw.. But there are webpages where you can print a template that should get me 98% of the way, the rest is just fine tuning. It's a time consuming process that I've delt with before when I did the exhaust manifold to the Kadett (stainless 1.5mm wall. I got sloppy on a couple of joints wich of course pulled the tubing out of my planned alignment..)

Nessism: Yes, the jig will help mostly when mittering, especially when tha plan is to build with lugs. A fillet brazing build would be more helped by a jig since there's nothing holding the pieces in the right place

duanedr: Great info! That experience is something I lack and was hoping that someone could share their's instead Are our numbers influenced any by the steel being heat treated or not? From what I have understood heat treated steel is stiffer than not heat treated, not just "harder"? I also interpet this as that the down tube is the "important" one of the front triangle?

Nessism

The heat treating does not affect stiffness. That's a misnomer that's been carrying on for decades; back to the 753 days at least.

It's my understanding that the most important tube in the front triangle is the down tube. My first frame used a long butt .8/.6 down tube and it was considerably stiffer than my subsequent builds using shorter butt .8/.5 tubing.

Andrew R Stewart 

If the frame is taken as a seperate from the rest of the bike device then I agree that the DT is the member which most determines torsional (pedaling loads) stiffness. But the frame is not ridden without other elements that add their flex. I've learned that the fork's steerer contributes a lot to the overall feel of a bikes stiffness. Also the more the load is over the rear wheel (like panniers or rack top stuff) the more the top tube comes into play. Andy

Doug Fattic 

It is true that it is more difficult to control the heat on thin wall tubing so the joint doesn't get too hot. But the other more challenging problem is applying even heat so the frame doesn't warp and get out of alignment. You need to use a pattern that minimizes warpedge while at the same time making sure the joint doesn't get too hot. Just to be clear, heat will bend a tube to the side that is being heated so you have to figure out how to correct for that. The easiest way is to bend the tube back after it has been brazed. Of course while that is possible, it isn't the best method. However thin wall heat treated tubing doesn't bend like old Reynolds 531 and it might buckle instead. The other option is to just leave it crooked.

Accurate frame alignment involves a number of factors that begins with good miters and includes placing the spots to keep them together in the plane of the frame. In addition you have to consider the sequence of how to braze the frame together. After you have taken those steps you need to follow a heating pattern that minimizes distortion when brazing each joint. These complications are why everyone is suggesting that you use heavier non-heat treated tubing that absorbs brazing mistakes better and can be bent back into alignment after each joint has cooled.