What makes a good design report?

Hey guys,

After 4 years on Baja, I still do not know the answer to this question. The feedback is pretty nonexistent (except in Montreal 2008, only one I can remember).

For example, we have a report from 2007 that got 49/50 in South Dakota and 45/75 in RIT. Another report got 41/50 in Illinois and 57/75 in Montreal (that was 2008). The reports were the same. So it's hard to pinpoint any good points about a design report.

So, I wanted to ask what are some of your tips and tricks that you know work for the design report.

Do the judges want:

• more pictures?
• more numbers?
• more testing?
• more comparison?

I'm guessing its a little of each and you have to find the right balance that fit on 10 pages.

Thanks guys

The judges like to see the "why's" behind your design... it isn't a sales or marketing document.  In talking with judges in the past they like to see pictures, numbers, FEA, etc. but they really thrive on validation of that analysis and testing.  You can do all the FEA in the world, but if your models aren't accurate who cares?  (yes this is hard to prove in a report rather than at the comp. when you are talking to the judge face to face) 

If I were a judge, I'd want to see safety and practicality behind designs. Nearly all machines have some potential disaster lurking in the background ready to pounce. The best protection is a large margin of safety.

The way I see it, the Design Report is not written a week before its due. It is an ongoing documentation of your work with topics like what's the engineering problem, what's your solution, what alternatives did you explore, what assumptions did you make in coming up with solutions, what potential failures did you see, did your insurance policies work? Always check up.

Learning to write technically helps. Use good engineering words and logical sentences. I have read reports from supposedly great designers that contained nothing but mambo jumbo and needless sensational publicity. Those went into my trash can.

Edited by RonGeorge - Mar/01/2011 at 3:46am

You're right. Depends what the part is and you have to use judgement. FOS of min 2.0 is necessary in general. I'm sure you wouldn't be penalized for using judgment.

But I also think doing extended static analysis misses the point and I will try to stray away from that. A lot of the parts fail by fatigue and not most of us can spend time or have the resources for fatigue analysis. Extended testing can yield that data to those who don't have the computer resources. For example, you can run the car everyday for 4 hours, for a straight month and you could get to high cycle fatigue regime in your rotating parts. I think showing that you did some kind of testing in real world conditions will always stand out in a Design Report.

RonGeorge wrote: Learning to write technically helps. Use good engineering words and logical sentences. I have read reports from supposedly great designers that contained nothing but mambo jumbo and needless sensational publicity. Those went into my trash can.

While you have to follow their format, but just putting some smart words on a page doesn't make for a successful report. In fact, this can sometimes make you look like more of an idiot.

That said, the biggest thing that I aim for in our design reports is not only saying why we did something but being able to justify why we did it. For example, it doesn't work to just say you designed something to have a factor of safety of 2. You have to look at the part and justify why a FOS of 2 will fly or not. Might be fine for a brake pedal, but probably won't work for a front knuckle.

Taking it a step further, for something like suspension, you have to justify why you decided to stick to the parameters you chose. Why is that going to work better than everyone elses? And how did this affect everything else? And after the suspension was fabricated, did the kinematics really turn out how they were supposed to?

I could ramble on and on, but think of a design report as the little kid that keeps asking "why?", even after you explain it to them.

Never got a good design report score. NEVER! You think we would have gotten lucky once in 6 years, but nope.  We just kept adding more and more to the appendixes in hopes that we covered it all.  By 2003 it resembled a dictionary. In 2004 I had enough and turned in a 20 page report with 5 pages of appendix.  Got nearly the same score.

 

Static judging also blows.  We would get a poor static judging of our suspension then win maneuverability.  We felt like protesting that crap and demanding they adjust our static suspension score based on our dynamic maneuverability win. 

The biggest problem with the design report is the overall inconsistency.  Save for the section on flotation and water propulsion our 2010 reports were nearly identical between SC and Rochester.  We were 16th at SC with 68.33 and 4th at Rochester with a 71 competing with nearly the same field.

I wrote both those reports and I wrote this years report and my advice would be to cover as much as you can with the minimal amount of explanation.  Try to include details where possible but focus more on how your car works together as a total.  Explain why everything on you car was designed with the overall goal in mind and how everything works together.  Also cover all your engineering bases.  Throw in some hand calcs in an appendix and a few graphs for suspension parameters.  Use figures to convey the information wherever possible.  Don't forget a list of symbols and try to reference some meaningful publications.  Explain what goal you designed toward but gloss over the intimate details of how you got there.

When you're limited to 10 pages you should probably be spending a day just working back through the report to find instances where you can use shorter words or phrases to convey the same information that way you can cram some more in.  I know when I read a design report other than my own it usually takes me an hour to make it through the whole thing just because of the sheer information overload. 

@jeib : You're on the right track with documentation. It's also pretty vital when practicing in industry, for more purposes than just for the act of documenting. Imagine you're the chief engineer for Transocean and you're leading a team to build the Deepwater Horizon. When the sh*t hit the ceiling, and the whole world was watching a catastrophe on TV, where will you go scrambling for all the documented events that happened during the project period in order to safeguard your reputation?

Another point I have to make is that without overly filling up the references section, cite bold claims. Otherwise, it is just nonsense. An example below. The snippet is from a Design Report of School X written 2 years back.

"COMPONENT DESIGN AND STRENGTH Solid Modeling and Analysis The vehicle was completely modeled in SolidWorks and analyzed using CosmosWorks. A complete vehicle assembly was created in SolidWorks. This complete assembly included the chassis, suspension, steering linkages, brakes, wheels and drive train. SolidWorks was used to simulate the suspension cycling from full droop to full bump. Using this software allows us to check for component clearances, toe and camber curves, as well as Constant Velocity (CV) joint plunge. CosmosWorks was used to execute the Finite Element Analysis (FEA) on the chassis as well as the suspension components. CosmosWorks analysis provided values of maximum stress and deflections that occur under any given loading condition. FEA was used as means to check hand calculations made on the components. It is critical that hand calculations are made because of the high error software packages are known to yield. These design techniques were used to assist in verifying the design’s strengths and make material selections. "

I'm not vouching for any software packages, or saying that the statements in bold are true or untrue. But when I read the report, I stop here, and tell myself - wow, it'd sure be nice to know where the author of the report gleaned this particular bit of information - that the rate of errors are high in FEA software. Furthermore, what gives them the right to feel that the "hand calculations" are trustworthy? In this particular report, there were no attachments with the calculations.

Bottomline, if you can't back up claims like this, don't bother wasting design report space for conjecture.


Edited by RonGeorge - Mar/02/2011 at 4:10pm

ErikHardy wrote: Its typically the user who makes the mistake in fea not the program. A few hand calculations should be done to compare magnitudes, not precise numbers.

Well, we got a 5th place finish in design report. Thanks for the help guys, really helped 

Akron 1998 to 2004 wrote: Never got a good design report score. NEVER! You think we would have gotten lucky once in 6 years, but nope.  We just kept adding more and more to the appendixes in hopes that we covered it all.  By 2003 it resembled a dictionary. In 2004 I had enough and turned in a 20 page report with 5 pages of appendix.  Got nearly the same score.

Meant to write something earlier, but better late than never. Although it's a lot fuzzier than it was 3 months ago.

Edited by JeremyB - Jul/16/2012 at 6:08pm