Quest frame flex tested

[billyk]

Hmmm ... I'm no engineer but I do believe the formula is right (I'm a physicist - fluid dynamics). Your G is for 7505-T6 aluminum, right?

Another solution would be to move the joint at node 2 up to right under the handlebar (reducing L to near-zero in your toolbox above). There would need to be another clamp at the top of the headset to hold that together. However, this would also lengthen the boom/extension tube, which would worsen its ability to withstand bending.

How much? How did you model the fact that the extension tube is clamped into the boom for 2-3 inches? And do you allow any flex in the joint at node 2?

I can't lower the handlebars without hitting my knees.

But, yes, I think it IS worth increasing the stiffness of the front end. My test rig (described at the top of this thread) shows significant bending under real-world high load. In fact I can see this by eye. That means part of the work I'm doing is going to bending metal rather than forward motion. The fact that this occurs precisely at the time when I'm working hardest - going uphill - makes it more important. It hardly matters for a nice cruise on flat ground, but climbing is where recumbents typically lose out to uprights, and the Cruzbikes have a potentially big advantage.

I don't have a Silvio or Vendetta to test but I'm also surprised to hear that the difference is only 15%. Lots of people on this forum ride both Qs and S/Vs, and they report much better hill-climbing.

It would be pretty easy to test the twist in the steering tube: just attach clamps top and bottom, with needles to nearly touch in the middle. Any relative motion of the needles would be visible.

Billy K

 

[cllsjd]

I realized I had made in error on saying it was 15% and I removed it.

The G is for aluminum in general. Depending on your source and type of aluminum it may vary some.

All joints in this beam model are perfectly rigid.

I've made a course model. The clamping of one tube in another tube is not modeled. I'd have to think about how even to do that and probably would need some detail dimensions. With two or three diameters of overlap my opinion, not fact, is that there isn't much flex in the joint.

At this point my calculation show you're wasting your time making the boom / extension tube stiffer until you've done something about the steering tube.

We've not touched on the flex in the handlebars.

 

[super slim]

For us mere Mortals, not Larry Oz, 100 watts is only 7 kg force (15.4 lb) on a pedal with a 175 mm crank rotating at 80 RPM. OR for Larry, 21 kg (46.2 lbs) pedal force for 300 watts
watts = kg*9.8*175/1000*rps*2*pi(3.146)

From my calculations 300 watts on a softrider with a light 100 kg (220 lbs) rider, will reach 9 mph (15 kph) riding up a 5% grade.

The 50 kg (110 lbs) pedal load at 80 rpm is equal to 630 watts for a 175 mm crank, so could only be maintained for a short time.

but this 100 watts creates a 25 Kg.f (55 lbs) chain pull for a 24 tooth granny chain ring or for 300 watts 75 Kg.f (165 lbs) chain pull.
Any larger chain ring will create a proportionally smaller chain pull.
This force is at approximately 60 degrees below the maximum pedal force direction.

So a lot of different forces acting on the Bottom bracket and its connections.
No wonder John T moved the clamps out to the external bearings on the Silvio!

 

[billyk]

The G is for aluminum in general. Depending on your source and type of aluminum it may vary some.

Well, here's the problem. Aluminum is hugely variable in strength and stiffness. Easily by factors of 3, 5 or more (link below to one such table). The Cruzbike frames are built of 7075-T6 (one of the 7xxx anyway). I've done a lot with soft aluminum (2024) which is very easy to work, roll, cold-form. This is an entirely different beast than 6xxx or 7xxx, which are aircraft grades and just not the same thing at all. Especially with the hard T6 temper used for Cruzbike frames.

I suspect the model results are going to come out quite different for different aluminum grades and tempers.

I'm also a little confused by your value for G = 3800000 psi. Which property is that (tensile, shear, yield)? But the tables I see online have no numbers that large, they're 2 zeroes smaller.

http://www.engineershandbook.com/Tables/aluminumprop.htm

Billy K

 

[cllsjd]

"Aluminun is hugely variable in strength and stiffness." You've hit on my pet peeve. Yes, aluminum does vary in strength. For a given cross section it absolutely, positively doesn't vary in stiffness. You can argue with me until the day you die and you'll never be right. Never!!! For a rod under axial tension K (stiffness) = AE/L. A = area E = modulus of elasticty L = length For aluminum E = 10x10^6 psi is the generally used value. For a beam in bending EI is the flexure rigidity or sometime thought of as the beam stiffness. I is the area moment of inertia. Again E is the modulus of elasticty and again 10X10^6 psi is the generally used value for aluminum.

My model results will not come out one bit different for any type of aluminum, because E or G vary little from one type of aluminum to the next. You can use the follow equation to relate E and G. v by the way is Poisson's ratio. I'd use .3 for aluminum.

(I stole this from wikipedia)

If you don't want to beleive me go look at MatWeb to find E. Pick any type of aluminum you care to. E will be about the same for all of them.

G is the shear modulus. The engineer tool box gives a values of 3.8 to 4.0 X 10^6 psi. Again it varies little from one type of aluminum to the next. It would appear that I have the correct number of zeros.

The reference you site is useless in this case as it only give strength values. I've made no arguements nor do I intend to offer an opinion as whether I think the frame is strong enough or not for the applied loads.

 

[cllsjd]

Super Slim;
I agree with you about my load. I went out in my driveway and had my wife helping me with measuring deflections. It was only after we were done did I realize what a huge gear (53-16 and 155 mm crankarms) that I was in. It was okay for what I was doing, but you're 100% right about not actually being able to ride in that gear.

 

[cllsjd]

I used the loads from Super Slim on my Quest beam model. The deflection at the bottom bracket was 3/16".

To satisfy my own curiosity, I added a couple of beam elements to the top of the steering tube to sort of simulate handlebars. The deflection of the bottom bracket increase by 1/16".

Below are the displacements at the nodes for these last two runs.

 

[cllsjd]

I made a foolish error. I assumed the steering tube was aluminum. I had it off this morning and it seemed heavy. A magnet verified it is steel. Time redo my numbers.

 

[cllsjd]

I hate to admit it, but I did something wrong. The measured displacements are just not matching up with the analysis. The analysis is actually saying that the displacement of the stock setup and a slivio type set up have roughly the same displacement. The measured displacement indicate the slivio type setup to have much less displacement.

I have found that with the slivio type set up it is much easier to keep the headset adjusted. That alone was worth the effort for me.