Dual 700
Dual 700C Quasi-Lowracer

A project by Warren Beauchamp

After building my bent tube chopper, I have been thinking about other bikes with bent tubes, like a dual 700C highracer. I have very long legs so it should be possible right?

This was the first design with the same riding position as the NoCom. It was decided that since this is to be a racing-only bike, it would be better to lay the seat back more.

This is the second design, with a more laid back 13 degree seat and a 6" seat height. Unfortunately, a 36" inseam is still not enough leg to clear the  wheel with the seat this low.
Design 3 has the same 13 degree seat, but a 12" seat height.
This looks like it will work. The next step is to see if I can bend the 2" x 0,049" tube.

I obtained the chrome moly 2" x 0,049" tube and spent several hours working on bending it. It is now more bent, but not bent enough, and I stripped the set screws on the roller bender 2" dies. I'll fix that and start bending again.

On other fronts, this dual 700C bike would be fun to build for my wife.
I spent many hours over several months trying to roll bent the 2" x .049" tube. It's about half way bent but there is no way it will get to the radius I need. Tom Schneider sent me a 2" x .035" tube, so I will try that next.

Top tube is the 2" x .049" tube that I spent 3 months trying to bend. Bottom tube is the 2" x .032" tube from Tom that I spent an hour or two bending today. That's a 700C wheel. Looks like one more bending session will get it to the correct radius.

Here's a quick shot of the scale drawing of the latest design. Crank overlap, 15 degree seat, 12 inch seat height. I now have the following parts to use with this bike:

  • XL Velokraft CF seat

  • Front / rear disk wheels

  • Steel forks for front / rear

  • Piles of bike parts from other projects.

Another half hour of bending and the tube is the right diameter. It's a bit ovalized but looks good. That seat looks small compared to those wheels and the tube.

I have been told I should make the seat completely laid back. This is what it would look like.

Here it is with a 10 degree seat. I think that's as laid back as I can go. Now it's time to to a 1X drawing.

I have most of the 1X drawing completed, except for the front fork. It appears that I will be looking through my hands and legs to see the road. I suppose that is what will make this fast.

Before the 1X drawing can be completed, I will need a fork. Forks have so many variation in stack height and rake that this one part can make a huge difference in the design of the front geometry.

I just bought this used road fork off ebay. It has a very low stack height, and should allow me to keep the planned 2" spacing between the tire and the frame.

I finished the 1X drawing and verified the fork would work.

Next step is to build the back half of the frameset. I am using an MTB fork for the rear wheel stays. Using the fork as shown in the picture would have had the dropouts pointing upward, not a great idea.  I un-brazed the rear dropouts, removed them,  reshaped the tabs, inserted them back into the dropout slots backwards, and rebrazed them.

I'll be using two side by side 1.5" x 0.32" tubes to connect the curved front frame to the rear. This will give a nice flat base for the seat to mount to.

I cut the head tube off of the MTB fork, cut slots in the tubes and inserted the fork.

The weather was nice so I jigged up the fork and bottom tube and then brazed them together. I didn't use the jig when I was re-installing the rear dropouts, which was stupid as it was of course it was not straight and I had to fix it. Fixing it usually involves some kludge and is never as nice as doing it right the first time.

In this picture I am checking the brazed part against the 1X drawing and eyeing it up with the from frame section. Looks good! That long thin tube shows what will be my eye line. I will be looking through the frame. Hopefully I will be able to get the handlebar low enough to keep my hands out of my line of sight.

Here you can see that by using the two side by side tubes, the brake bridge is still exposed. This will allow me to mount a brake easily rather than making a custom brake mount bridge.

This joint took about 3 rods of braze to finish. I have used almost all of the 5 pound tube of rods over the past 8 years or so.

I cut the fishmouth into the bottom tubes this evening with a 2" hole saw in my drill press. I had to do it twice as the first time it was 10 degrees off. Measure twice and cut once and all that jive. Because of this, the wheelbase is now about 1/2" shorter than drawn. Good thing I had a little room to spare.

It's pretty close now and I will use a hand file or the grinder to finish it up once I get the whole frame jigged up and ready to braze.

Every time I build a bike I rediscover how many different ways there are to mess up. Good thing I am not a perfectionist.

I jigged up the bike and brazed the bent tube to the bottom tubes. It looks pretty good. This picture shows it with the seat resting on the bottom tubes and the reference rear wheel. The seat will slide forward a couple inches when I notch it to fit into the bent tube.

I had a productive day today. The front fork was widened using a conduit bender. The rear drive wheel fits into the front fork nicely now.

Before cutting the head tube hole, a lot of time was spent measuring and eyeballing frame while mounted in the drill press.

Here's the head tube temporarily mounted in the frame for me to eyeball it.

I jigged up the frame on my frame jig and then brazed in the head tube. It looks like I may need to take a little of the rake out of the fork.

I have a bottom bracket shell, bottom bracket, and some chain on order. I did not order any fancy parts, as this will not be a lightweight bike. I am hoping it will be under 30 lbs.


  • Cut the hole for the bottom bracket and brazed it in place.

  • Mounted old Sugino 52T chainring into my 160mm DaVinci cranks.

  • Took some of the rake out of the front fork with my trusty conduit bender. The wheel is better centered in the frame arc now.

  • Brazed on the power side idler bracket

  • Cut a big slot out of the Velokraft seat so it slides forward until I am intimate with the down tube.

  • Sat on the bike to check to see if I could actually reach the pedals (yes!).

  • Check to see if I will have heal clearance issues with the derailleur (no).

  • Installed headset. Wow was that tight, I hope I didn't trash it.

Built up some adjustable handlebars. I also added a big tube on the front brake to space it out far enough so that it doesn't contact the drive side chain.

Took the bike outside for it's first test ride. Can I ride it? Yes! It took several tries but I got the hang of it and rode it up and down the street. I had to use the flintstone brakes to stop it so I didn't try to go very fast. Now it's time to cap tubes, add some bracing under the seat, and add a head rest. The bike weighs about 27 lbs in its present configuration.

Capped the tubes and added a gusset. The bike was a little bouncy before I did that, and is less so now.

Seat height is 12". BB height is 19"..

Ready for a short test. I installed the power meter front wheel on the bike and went to a section of the bike trail that is flat for testing.

Here are the results of the testing (click on picture for a bigger view). I tried to hold the wattage around 200 watts, but the trail was very busy so I had to let off a few times. Also, the flat section is only about 3/4 mile long, so  had to turn around pretty quickly (big power dip in middle of plot). It would have been easier to test at the velodrome, but that is a long drive.

Results seem to say that the bike goes 25 MPH at 200 watts, without the disk wheel. I'll try again when I add the disk wheel and see what kind of difference it makes.

I have to look through a lot of stuff to see the road...

I figured out how to use a back caliper brake as a front brake. Back brakes have shorter mounting bolts than front brakes, and typically they are not long enough to go all the way through the fork crown. The trick? Use an old canti brake boss (left) and file off the area that you would wrench it on with (right).

Thread it onto the caliper brake mounting bolt from the back. Then you can use a 6mm nut on the back of the fork to fasten it.  Now you have twice as many brakes in your brake box as you thought!

I added plastic covers to my trispoke CF wheel yesterday. I didn't have any ABS, so I used some Lexan sheet that I painted black. It's probably a bit on the thick side, and the wheel now feels heavy, but on the plus side it will be bullet proof.

I attached it with double sided carpet tape. This stuff is incredibly sticky and stretchy goo. It's all sticky stuff, with no backing. It was a challenge just to get it off the roll and onto the wheel.

I brazed the T-nuts in to fasten the seat firmly and ran the rear brake cable through the frame. I noticed that the lexan wheel covers had popped off of the wheel while I was on vacation. The lexan sheet had been rolled in a hot attic for several years, so it really wants to be curved. I threw them out on some hot asphalt to see if they would relax a bit, but that did not seem to do much. The tape seems to be sticking to the wheel fine, but not as well to the lexan. I roughed up the lexan a bit and added more tape. Hopefully it will stay stuck.

I'm hoping to take another (longer) test ride this weekend after I have the disk wheels installed, and maybe do a coast down comparison with the nocom.

When you build a FWD recumbent, by default, the derailleur cable is pointed 180 degrees in the wrong direction. To fix this, you have to make some parts. I made a cable stop that bolts onto the cable clamp bolt out of a cable stop braze-on, a washer, and a 5mm nut.

Brazed together, works nice.

Here's a shot of the D7CLR vs the NoCom. I did a coast down test comparing these two bikes to see if it was really worthwhile building this new bike.

D7CLR - 29lb
NoCom - 27lb + 2 lb clay in the seat.
Coast down length - .64 mi
Max speed - 25 MPH
Tires - 110 psi.
700C tires all Vittoria Corsa CX.
Front NoCom tire - Schwalbe Ultremo

Result - Amazingly similar, but D7CLR coasted 57 ft further than the NoCom. 57 / 3379 = 1.6% faster. At 26 MPH, that's potentially about .4 MPH faster.

I cut off the excess tubing at the bottom bracket and brazed in some tubing caps to make it pretty. The frame is done! Next it was time to take it all back apart to paint it. A coat of shiny black paint on it out of a rattle can was nothing fancy, but it looks fine. The lexan disk popped off again. I am going to try relaxing the bend a bit with a heat gun. I think it will be ready to race in Northbrook. 

Front view.


Northbrook velodrome races Pictures by Courtney Shov-lynn.

Last weekend at the Northbrook races was the first real test of the D7CLR. The rear wheel covers promptly fell off when I set the bike out in the sun, so I raced the bike with a rear tri-spoke wheel and front wheel disk. The 50 lap stock race was directly after the 100 lap 'liner race and I was pretty toasted, so it took about 20 laps before I felt like I could push. Even so, average speed from the HPRA timing system was 25.72 MPH. Speeds were higher after about 20 laps when I was actually able to push.

The biggest issue with the D7CLR was that I was looking through a bunch of junk to see the road. This is ok for a time trial, but not so good when I am trying to draft. The NoCom is so much nicer in that respect because I can see the road and it is easy to draft. On Sunday I raced the NoCom and averaged 27.5 MPH. Based on that the speeds seem pretty comparable.

I did get consensus that I can cut down the head tube a couple inches, that will help a bit with the vision. There are some other tweaks to make and then I will try to race it again.

I raced the D7CLR at the N. Manchester HPRA races last weekend, and wore the Kevlar shorts and sleeves to prevent road rash in case I crashed in the tight corners. Picture by Ben Cooper.

I took the disk wheel off of the Cuda-W streamliner. It worked pretty well but I definitely need to reinforce the rear stays. I won the lap races and it helped that Mike, Sean, and Dennis stayed home. Also I still need to cut down the head tube and adjust the handlebars up to brings my arms into a more aero position.

If you look closely you can see the rear brake that fell off with a couple laps to go on the Sunday lap race. I had to stop and quickly remove remove the brake bolt that was rubbing on the tire before Ben unlapped himself and won the race. Note to self - make sure brake is more than finger tight before racing...

I hade several changes to this bike over the winter. I shortened the head tube to bring the handlebars down a bit, and rebuilt the rear wheel covers. They looked great in the winter, but now that its warmer they are a little wavy. The expansion coefficient of the PETG wheel covers must be a lot different then the CF and aluminum wheel that is under them.

I added 3 layers of CF to the rear stays to reduce the bounciness. Its much better now. I also fixed the rear brake mount so the brake doesn't fall off again...
Here's a close-up of the FWD pulleys and the front brake showing the mounting to avoid the FWD power side chain.

I'll add some plastic under the seat to clean up the airflow there before the Northbrook-Kenosha velodrome races this year.

I raced the D7CLR after racing the streamliner on the 100 lap Northbrook velodome race, so I was not exactly fresh for the 50 lap unfaired race. The bike performed well, but no better than the NoCom.

I decided it is too squirrelly with the front wheel disk. Riding it with my deep dish wheel seems fine. Also I really hate looking through the headset to see the road. Because it's not fun to race it, and it's no faster than the NoCom, I will be de-radicalizing this bike to make it more road friendly. This will involve raising the seat back to make it high enough to see over the headset, changing the handlebars to be higher and running something other than disk wheels. I think it will still be fast, and also more fun to ride.

Stay tuned!

Today I hacked the back and front of the frame apart.

Here the highly scientific test that I used to determine how high I had to raise the seat to ensure I would be able to see the road far ahead without looking through the headset. This indicated I will need to raise the seat 15 degrees to make 25 degrees total which is coincidentally the same seat angle as my NoCom and CF highracer.

Rather than just raise the seat back to get the correct seat angle I am reconstructing the rear half of the frame. This will keep the bottom tubes embedded in the seat and allow me to bolt the seat directly to the frame tubes. Here's the new scale drawing. The seat is raised to 25 degrees, the rear stays are rotated 10 degrees, and the handle bars are rotated up 20 degrees. The wheelbase will be several inches shorter. This looks remarkably similar to my original concept drawing for this bike...
I did some more hacking and laid out the 1X drawing. It looks like the seat was really at 5 degrees before hacking, and that it will be 20 degrees after reconstruction. Probably.

Here are all the parts laid out on the 1X drawing. Wheelbase will now be 51 inches with the rear wheel tucked in tight behind the seat..

After much cutting and jigging and brazing, the bike is more or less back together. Due to a lack of double and triple checking, it ended up with a bit less trail and a bit lower BB than designed.

Seat - 16"
BB - 20.5"
Trail - around 1"
Wheelbase - probably 52"

I also hacked off the handlebars and moved them 10 degrees up. A quick test ride showed that at low speed it still handles fine.

I need to cap the tubes, add gussets and fasten the seat with something more than a couple sheet metal screws before taking it for a real test ride. Also now the 160mm pedals are just a hair more of a reach than I would like. If it looks like it will be a problem I will just go to 155 or 150mm pedals.
I cut some slots into the Zote foam seat pad for a little cooling and for looks. It's pretty easy to cut the slots with a router.

I probably would not use yellow for a seat pad again as it shows every bit of dirt.

I decided to reinforce and stiffen the joints with CF rather than more brazing. Here the joint is wrapped in a few layers of epoxy soaked CF, then wrapped in plastic wrap. It's all wrapped in electrical tape to compress it and form a good bond with the steel. I used some chunks of foam to press the CF down into the areas that the tape would not have been pressing on the CF.


On the joint where the down tube attaches to the bottom tubes, I built a box made from CF using an epoxy / cotton flox mixture to stick it together. I will then cover the whole joint with a couple layers of CF. The box is there to cover a tube I had added to strengthen the joint, and to fill in the area under the seat pan for a little extra aero.
I finished up reinforcing the front joint shown above by wrapping it with a couple layers of CF. As before I covered the whole thing in cling wrap, added foam to strategic locations, and wrapped it all with a bunch of electrical tape. 
 Here's the back joint after giving it a finish coating of epoxy.

They turned out well and the bike has lost a lot of the bounciness that it had in its previous incarnation. 

I weighed the bike and it still weighs 29 lbs.

Next was to fix the shifting. My current solution only got into 6 of the 9 gears. This is a common problem when mixing Shimano shifters and SRAM derailleurs.

I had to file the Shimano index shifter down so it would have a longer throw and work with my 9spd SRAM derailleur and gears. You can see in the picture that there is now a large gap between the shifter handle and the shifter body. The shifting is working properly now.

Also you can see a bit of the handler that is wrapped in innertube as a substitute for traditional handlebar tape. I never have to buy handlebar tape again!

The bike is almost ready for racing. Now I just need to touch up the paint and add some number panels!

Dave Mendea took a bunch of nice pictures at the 2015 Waterford, MI HPRA races.

Here's a picture of the bike ready to race the one hour race.

This picture was taken during the one hour race. I finished 4th overall and my speed was 24.36MPH for the hour. This was great considering the curves and the hill on the course. I was about .3 MPH slower than the 2nd and 3rd place racers, and about 1.5  MPH slower than the winner.
This picture is of me doing the hill climb. That is my go fast face. :)

The bike performed very well and I am happy with the ergonomics as now I can see over the headset! It seems to be about the same speed as the NoCom. Also it is more comfortable due to the large wheels and compliant frame.

I do need to make sure to wear different shoes, as these have a knob on the heel to tighten the shoe. The knob occasionally hit the derailleur which made the shoe loose. I had to unclip and re-tighten the shoe while riding several times during the hour race.

The only other issue I had was during the hill climb. While cranking hard at high RPM I had front wheel hop which would cause the bike to lose traction. I had to back off several times which hurt my hill climb / coast down performance. I think that a couple layers of CF on the down tube will fix that.



I made a video of taking the bike around the Waterford track, using the GoPro helicopter mount helmet cam. The mount was made from a long stainless tube with a hole drilled in the middle. A long carriage bolt goes though an old helmet from the inside, and a bunch of washers provide the bearing surface so it will spin. The camera is mounted on one end of the tube and on the other end is a counter-weight. Its a bit heavy in its present form and would be easier to use with a lighter tube. Not that I can use  it when anyone else is close anyway...

The camera moves in several ways.
* Inertia - when going around a corner it wants to stay in the same place.
* Wind - The camera end has the largest surface area, so it wants to move to the back.
* By hand - I can spin the camera around when I want to give a quick 360 view.

Picture by Larry Oslund

The bike performed well at the At the Northbrook 100 races. The engine did ok too.

After several seasons of tweaking and testing at races, I think it is at least as fast as the NoCom.

Picture by Tony Levand

After several years of designing, building and racing this bike I have decided to build a carbon fiber version.

Next - The Carbon Fiber D7LR

In 2016 I sold the D7CLR to John from Phoenix. He has been modifying it and recently added a bunch of fairing. Looks fast!


Back to the HPV Projects