Difference between revisions of "Shawn Pierce's Course Project"

From GICL Wiki
Jump to: navigation, search
Line 119: Line 119:
  
  
 +
'''Side View of the Manticore'''
  
 
[[Image:SideViewManticore.JPG]]
 
[[Image:SideViewManticore.JPG]]
 +
 +
 +
'''Front View'''
  
 
[[Image:FrontViewManticore.JPG]]
 
[[Image:FrontViewManticore.JPG]]
 +
 +
 +
'''Top View'''
  
 
[[Image:TopViewManticore.JPG]]
 
[[Image:TopViewManticore.JPG]]
 +
 +
 +
'''Bottom View'''
  
 
[[Image:BottomViewManticore.JPG]]
 
[[Image:BottomViewManticore.JPG]]
 +
 +
 +
'''Diagonal View'''
  
 
[[Image:DiagonalViewManticore.JPG]]
 
[[Image:DiagonalViewManticore.JPG]]
Line 137: Line 150:
 
The virtual model is nearing completion, more on this soon.
 
The virtual model is nearing completion, more on this soon.
  
As of Thursday Nov 16th at 4:02pm the manticore lives.
+
As of Thursday Nov 16th at 4:02pm the manticore lives. :)
 +
 
  
 
'''A Side View of the Beast'''
 
'''A Side View of the Beast'''

Revision as of 16:12, 16 November 2006



Contents

The Robot's Design

By: Britt Martin

I have been calling my robot "The Manticore". It gets its name from the fearsome beast which has the body of a lion, the tail of a scorpion, and the wings of a bat (only much larger). While my robot does not have wings or a poisonous tail, it will have a stride that I imagine is much like that of the Manticore.


The Manticore

Here are some of the photos I have taken so far:

An Early View of the Frame

Early frame.JPG

This shows the two motor setup I am using. One thing that I have already noticed, as I uploaded this image, is that I will need another gear to reverse the motion of the hind legs.


A Top-down View of the Body

Body1.JPG

Not too shabby...


A Diagonal View of the Body

Body2.JPG

Getting there...




Timeline

Week 1

Researching possible robot designs, building with LEGOS™, and taking pictures.


Week 2

The current robot consists of a body which makes use of two motors that will power two legs each. I am using a locomotive type of motion to allow the legs to walk. Early on I realized that I will have to stagger the rate at which each leg touches the ground, so that it will be able to walk in stead of stand in place and gyrate its body.

I am working on powering the motors using an older Lego system that predates the Mindstorm kits, but still uses the same connectors.


Week 3

This week I am hoping to finish up the robot by making a few adjustments on the body, and then by adding its legs.

I also hope to begin assembling a CAD model as soon as the physical assembly is complete.


Week 4

So far I have tried out 3 CAD environments.


Bently Microstation™

The first of the three was Bently Microstation. I found this piece of software particularly hard to get used to as far as its user interface goes. I was not able to load the step files for the lego pieces in Microstation (not to say that it cannot be done, but after about an hour of messing around I decided I would try something else).


UGS I-deas™

The next piece of software I sampled was UGS I-deas. This program seemed to be a step up from Microstation. It actually had a mechanism for moving, rotating, and translating objects. However, to move an object you have to go through red-tape-heaven. Left click on the object, right click on the object, select DONE then go to the move toolbox, then select which type of move you want to do, then do the move. Simple huh? Well it really breaks your train of thought quite thuroughly. And after all the hassle, you are not in some sort of "move mode". Nope, if you want to move a part a second time, you have to go right back to the beginning. This goes for each of the other operations that I tried. There may be a much easier way to do things in this program, but it is not made apparent.

Megan let me try modeling with her already imported lego setup. It was fairly easy to bring legos out of their "bin" and into the assembly area. There is, although, a catch. You take the lego you want out of the "bin" and put it on the viewable area of the screen. Now, to your surprize/dismay, it does not exist in the "bin". So, say you made a robot with more than one of the same piece, you now have to make copies of these parts. This is okay, but it is just confusing as to why the designers of this software would make so many choices that confound the same process that they're trying to facilitate.


MCS Adams™

Adams, I thought, would be the solution to all of my problems. It has a slightly more polished interface than that of I-deas. It was able to load STEP (.stp) models without too much trouble. You must first import the model, and then navigate a hierarchy of the model to select the specific "part" that you want. That can be tricky, but luckily Mike outlines the process quite nicely here.

After loading a model into the Adams viewer, you can rotate, translate or zoom the view (using the "r", "t", and "z" keys and a mouse movement). I was unable to actually rotate, translate or scale the actual objects (scaling being of very little importance for our purposes). I was also unable to load multiple parts at the same time.


Now, call me crazy but...

I have been thinking a lot about any of the good modeling experiences that I've had. The one that seems to stand out in my mind is 3D modeling using Maya. Now, you might be thinking to yourself, "But Shawn, Maya is for making 3D models to be animated, and not so much for part assembly or production." To which I would say, "You're absolutely right, but, is there, by some off chance a way to do a good physical simulation of a model?" Surprisingly, there is. It is called "rigid body dynamics". In Maya, which, from what I've read can import STEP models, you can denote that a model, or part of a model, is a rigid body. From there you can assign a gravity field to the model, and much more.

I found a tutorial on this, and I think it might be of some use. Domino Effect

Also in the tutorial, there is this very compelling video. Domino Video


Week 5

I have decided to use Adams to model the manticore using the primitive 3D shapes that are offered in their system instead of the Lego parts. I am reading up about Adams and what this tool is capable of in terms of simulating a model built in it.


Week 6

I am currently working on finishing up the physical model, and getting a good design for my virtual model.


Week 7

I finished the physical model of my robot. I made a few significant changes to the structure, and also attached the legs. One of the most important changes, and also the simplest to implement, was that I had to make each of the legs step in the same direction. My previous model was based on a pulling motion to drag the robot, and each of the legs would end up dragging to their side. This would make it so that no net distance would ever be traveled. I made it so that it is now more of a stepping motion.

Another change that I made to it was to reduce the weight of the robot by simplifying the design. So that the motors are able to power two legs each, I also made the step size a little bit shorter.

Here are some of the pictures of the manticore:


Side View of the Manticore

SideViewManticore.JPG


Front View

FrontViewManticore.JPG


Top View

TopViewManticore.JPG


Bottom View

BottomViewManticore.JPG


Diagonal View

DiagonalViewManticore.JPG



Week 8

The virtual model is nearing completion, more on this soon.

As of Thursday Nov 16th at 4:02pm the manticore lives. :)


A Side View of the Beast

Media:TheManticore_SideView.avi

A Front View

Media:TheManticore_FrontView.avi

A Top View

Media:TheManticore_TopView.avi

A Diagonal View

Media:TheManticore_Diagonal.avi


This is a zip file with the manticore model file for Adams™(.bin)

Media:TheManticore_Model.zip


Interesting Links

Snake robots

Samuel L. Jackson's finest

Interesting Lego Modeling Program