-- MarkCutkosky - 17 Jun 2004

From: "Emily Joyce Ma" 
To: "Arthur Joseph McClung" 
Cc: "William R Provancher" ; "Dana Macy Ung"
; "Ji Lee" 
Sent: Wednesday, June 16, 2004 9:37 AM
Subject: BDML test track braindump

Hi Trey, Will, Dana and Ji

I'm really really sorry - been totally wrapped up taking care of the 'rents while they are here. Alas! I'm going to come in at least a little bit again once the undergrads start going for their first two weeks. (Going to China from June 18-25 an then back from June 25- July 12 to work on another hanging project . . . so I'll be in and out . . . until I start summer work).%ENDCOMMENT%

---

A. LEG AND FOOT TEST PLANS

TEST PROCEDURE

- I've put on Chewie a copy of the test plan that we used to run the first two sets of trials using step test - one just for kicks that Moto and I did way back in April and the other that Dana, Alan and I did recently with the accelerometer stuff (which I still have yet to process).

//chewie/Emily/Test Track/Preliminary Test Track Experiment v2.doc

PARAMETERS

- I was basically varying 'steptest_time_offset' and 'steptest_duration' to get the attachment and detachment at different times.

SPECIMEN ATTACHMENT

- I cut an extra 'lower limb' out of 1/4" acrylic that's in the box that we can use to screw Will's foot to as well as other future feet. The reason why I lasercammed an extra few of these is because I really mucked up the one that's on there right now - in the sense that I attached a piece of clear circular acrylic as a backbone to the current foot that's on there to prevent it from bending out of the plane.

N.B. The initial parameters are really sensitive to the following things unfortunately as Alan and I found fiddling with the system a bizzilion times to get things to sort of match up to the original trial runs.

• How far the entire test track is mounted away from the wall (SangBae sometimes changes this so I think we should mark it). Z-axis offset.
• That little piece of metal with the earplug in it makes a huge difference!!! Geezus. Sometimes when we jolt the test track and the support string pulls upwards, the piece of metal gets deformed and again, this changes the Z-axis damping and offset for the trial a little.
• Claws get bent. Uh, they are really sharp as you probably have found.

B. FRICTION TESTS

From talking with Will, we were going to do this by running only a portion of the trajectory, i.e. from just before attachment to just after detachment with a 35A foot (or maybe 10A foot) on glass - which is wrapped in brown paper in one of the boxes on Sangbae's table). I didn't get around to doing this but it will probably require the following:

1. Remove current foot and mount one of those flat speciments from Will with double stick tape or something. You might want to add some side 'rails' to the structure to keep it from torquing side to side or secure it really really well with hot glue. 2. Put glass on the force plate. 3. Come up with the initial step_test parameters that will bring the foot really close to the surface, do attachment-stroke-detachment, and lift off a little from the surface.

Changes:

• steptest_time_offset ~0.00 since start at attach
• steptest_duration ~final_time since only want stroke
• calib_offset0 Want start at top of attachment
• calib_offset3 Want start right near surface
• steptest_final_time ~duration since only want stroke
• steptest_amplitude Very little to bring it onto surface

I also built a lego getup (do you see it near Princess Leia) to do basic static tests. You just screw this onto the test track and it constrains any motion in the x and z directions so you can bring the structure with the foot on it close to the plate (i.e. foot touching) just add weight to it (or tie it to the motor) until the foot slips. You should be able to determine the coefficient of static friction from the that since the force plate will give you a good reading for the normal force and lateral force at the point of slippage.

C. THE UNDERGRADS

Are TOTALLY READY TO RUN TRIALS!!! Dana and Ji did one entire trial themselves taking data for Alan's accelerometer. The only thing that's hard is making sure things are setup properly - in terms of, are we seeing forces that we expect. They know how to set the test track up, how to run bailorgana, how to run LV on princessleia. I think they are starting on JUNE 22.

Okay, shoot me a email if there's something I can clarify!!

Have a good start to the summer.

Em