Page 119 of 124

Re: Elevated Bridge B/C

Posted: May 25th, 2010, 2:28 pm
by eta150
Exactly

Re: Elevated Bridge B/C

Posted: May 25th, 2010, 2:52 pm
by nejanimb
We would've gotten second with that scoring ;-).

Interesting concept though. I kind of like the incentive to go higher on weight held... although that also leaves a lot more room for problems if your structure collapses early. I guess that's a good thing - encourages you to be safe and use sound engineering practices - real structures ought to be designed to never fail! I definitely wouldn't like going back to where you'd get second-tiered if you didn't hold full load though. Maybe there could be a minimum load for tier one that was different than the max load? I sort of hope they wouldn't call that new version "efficiency" though - I like how, with the current name, efficiency actually is what it is: a bridge that scored 2000 actually supported two thousand times its own mass.

Re: Elevated Bridge B/C

Posted: May 25th, 2010, 2:57 pm
by Littleboy
I will finally break 1500 efficency. It would be easy for most of us to get well over 10 mil

Re: Elevated Bridge B/C

Posted: May 25th, 2010, 3:38 pm
by lllazar
Man that sounds pretty cool...ive always built my prototype design and tested it on as much sand as it can hold, and then worked down/up on the mass of the actual bridge...

Re: Elevated Bridge B/C

Posted: May 25th, 2010, 5:00 pm
by rjm
Interesting twist on the scoring for Towers. I was unable to attend the meetings on Sunday to hear the rationale or comment on the extent to which this idea will weight the scoring. Hopefully, whoever came up with this one thought it through and tested some scenarios. Up through the 2001 season, structure events were scored as structural efficiency + 10% of load held. The 10% was a bonus to encourage competitors to hold more weight, but it was not so overwhelming that it discouraged teams from taking risks with lighter structures. For a given bridge mass, the scoring increased linearly with greater mass held; across a range of bridge masses, it was roughly a hyberbola. That system encouraged higher masses held while still allowing lighter bridges with lower masses held a decent chance of winning.

Scoring by using the mass held squared is not analogous to the structural design of beams, cantilevers, or bridges loaded with concentrated loads. Strength formulas for beams or bridges resisting concentrated loads use the load as a direct multiple; formulas using distributed loads use the square of the distributed load, but that's not in play here. Basic column buckling formulas such as Euler's formula use the first power of the load to predict a buckling threshold as well, and while other formulas such as Johnson's parabola or the secant formula are not first-order, they are not similar to using the square of the load either.

Aside from an arbitrary weighting scheme for scoring, was this formula intended to teach students anything about structural design or efficiency? I suspect it will simply squeeze out all of the low-load structures. It reminds me of the days when breaking at less than 15 Kg was second tiered.

Bob Monetza
Grand Haven, MI

Re: Elevated Bridge B/C

Posted: May 25th, 2010, 6:20 pm
by lllazar
Interesting twist on the scoring for Towers. I was unable to attend the meetings on Sunday to hear the rationale or comment on the extent to which this idea will weight the scoring. Hopefully, whoever came up with this one thought it through and tested some scenarios. Up through the 2001 season, structure events were scored as structural efficiency + 10% of load held. The 10% was a bonus to encourage competitors to hold more weight, but it was not so overwhelming that it discouraged teams from taking risks with lighter structures. For a given bridge mass, the scoring increased linearly with greater mass held; across a range of bridge masses, it was roughly a hyberbola. That system encouraged higher masses held while still allowing lighter bridges with lower masses held a decent chance of winning.

Scoring by using the mass held squared is not analogous to the structural design of beams, cantilevers, or bridges loaded with concentrated loads. Strength formulas for beams or bridges resisting concentrated loads use the load as a direct multiple; formulas using distributed loads use the square of the distributed load, but that's not in play here. Basic column buckling formulas such as Euler's formula use the first power of the load to predict a buckling threshold as well, and while other formulas such as Johnson's parabola or the secant formula are not first-order, they are not similar to using the square of the load either.

Aside from an arbitrary weighting scheme for scoring, was this formula intended to teach students anything about structural design or efficiency? I suspect it will simply squeeze out all of the low-load structures. It reminds me of the days when breaking at less than 15 Kg was second tiered.

Bob Monetza
Grand Haven, MI

There were days when holding less than 15kg was second tier? Who came up with that? That basically takes all risk out of the equation...and even one bad piece of wood could screw up your whole structure. That would be frustrating...

Re: Elevated Bridge B/C

Posted: May 25th, 2010, 9:58 pm
by andrewwski
Yes - through 2006 I believe.

I agree with rjm's thinking. I'm curious as to what the reasoning behind the new scoring method would be. Obviously to give higher rankings to those who hold more weight, but why?

There will be a number of situations where a higher score will be given to a less efficient bridge. In this case, what is the goal of the event? To be the most efficient, or to hold the most weight? Why does it make sense for a 15 g bridge holding 15 kg (efficiency of 1000) to have a higher score than a 3 g bridge holding 6 kg (efficiency of 2000)? This seems to put too great an emphasis on weight held, with efficiency being secondary.

If this is the goal, then so be it, but I fail to see how it is a good goal. What is to be learned, from the actual design, from this system? In the above example, the 15 g bridge can be made without much analysis, just a lot of trial and error, while the 6 g bridge obviously requires much more attention to the details and workings of the design. Yet the 15 g bridge places higher.

Rjm's reasoning makes a lot of sense. This new method seems to have an arbitrary fit with what most would probably agree the purpose behind the event should be. I do not see how such an alteration to the method of scoring makes any sense.

Re: Elevated Bridge B/C

Posted: May 26th, 2010, 5:12 am
by Balsa Man
Interesting discussion on the implications of a new scoring approach. rjm's discussion of the mathmatical relationships involved, and the relationship (or lack thereof) to efficiency suggested by the new scoring approach made a lot of sense.
It can be difficult, though, to grasp the implications of.....a description of how numbers would be handled, so I "ran the numbers". Interesting and useful excercise, and I have to say, I was a bit surprised at the result.

The first part of the table below is the "new" scoring approach; with weight held running from 15kg down to 14kg, and weight of structure running up from 4.5gr to 5.4gr. The part below it is for the old/current, "straight efficiency." In each portion, 4.5gr carrying 15kg is set at 1- i.e., 100%, and the rest of the table is the score as a percent of the 4.5 gr carrying 15kg score. Sorry, but can't get the top line- the structure weights fm 4.5 to 5.4 - to align with columns

So, what does it mean/say?; what's "the bottom line"?

As jander14 indicated when he passed this on, compared to current scoring, it, indeed, gives a bit of emphasis to weight carried; not a lot; certainly nothing radical; it just builds in a little bit of bias. If you extend the weight held down to 10kg, the "fall-off' in score is more dramatic in the new scenario; a 4.5gr structure carrying only 10gr scores 44.44% of one carrying 15kg under the new, and 66.67% under the old.

Held 4.5 4.6 4.7 4.8 4.9 5 5.1 5.2 5.3 5.4
15 100.00% 97.83% 95.74% 93.75% 91.84% 90.00% 88.24% 86.54% 84.91% 83.33%
14.9 98.67% 96.53% 94.47% 92.50% 90.62% 88.80% 87.06% 85.39% 83.78% 82.23%
14.8 97.35% 95.23% 93.21% 91.27% 89.40% 87.62% 85.90% 84.25% 82.66% 81.13%
14.7 96.04% 93.95% 91.95% 90.04% 88.20% 86.44% 84.74% 83.11% 81.54% 80.03%
14.6 94.74% 92.68% 90.71% 88.82% 87.00% 85.26% 83.59% 81.98% 80.44% 78.95%
14.5 93.44% 91.41% 89.47% 87.60% 85.82% 84.10% 82.45% 80.87% 79.34% 77.87%
14.4 92.16% 90.16% 88.24% 86.40% 84.64% 82.94% 81.32% 79.75% 78.25% 76.80%
14.3 90.88% 88.91% 87.02% 85.20% 83.47% 81.80% 80.19% 78.65% 77.17% 75.74%
14.2 89.62% 87.67% 85.80% 84.02% 82.30% 80.66% 79.07% 77.55% 76.09% 74.68%
14.1 88.36% 86.44% 84.60% 82.84% 81.15% 79.52% 77.96% 76.47% 75.02% 73.63%
14 87.11% 85.22% 83.40% 81.67% 80.00% 78.40% 76.86% 75.38% 73.96% 72.59%

Held 4.5 4.6 4.7 4.8 4.9 5.0 5.1 5.2 5.3 5.4
15 100.00% 97.83% 95.74% 93.75% 91.84% 90.00% 88.24% 86.54% 84.91% 83.33%
14.9 99.33% 97.17% 95.11% 93.13% 91.22% 89.40% 87.65% 85.96% 84.34% 82.78%
14.8 98.67% 96.52% 94.47% 92.50% 90.61% 88.80% 87.06% 85.38% 83.77% 82.22%
14.7 98.00% 95.87% 93.83% 91.88% 90.00% 88.20% 86.47% 84.81% 83.21% 81.67%
14.6 97.33% 95.22% 93.19% 91.25% 89.39% 87.60% 85.88% 84.23% 82.64% 81.11%
14.5 96.67% 94.57% 92.55% 90.63% 88.78% 87.00% 85.29% 83.65% 82.08% 80.56%
14.4 96.00% 93.91% 91.91% 90.00% 88.16% 86.40% 84.71% 83.08% 81.51% 80.00%
14.3 95.33% 93.26% 91.28% 89.38% 87.55% 85.80% 84.12% 82.50% 80.94% 79.44%
14.2 94.67% 92.61% 90.64% 88.75% 86.94% 85.20% 83.53% 81.92% 80.38% 78.89%
14.1 94.00% 91.96% 90.00% 88.13% 86.33% 84.60% 82.94% 81.35% 79.81% 78.33%
14 93.33% 91.30% 89.36% 87.50% 85.71% 84.00% 82.35% 80.77% 79.25% 77.78%

Re: Elevated Bridge B/C

Posted: May 26th, 2010, 7:20 am
by packer-backer91
The new rules make for a more real cross over into a real life application. So if 15kg is your max load that can be expected, it does make sense in a real world application to reward a team for not meeting the requirement. So because the a theoretical bridge weights 1g and holds 7kg of the 15 allowed give a score of 7000, the rule change makes teams go back to a real life situation when building. I think was that the goal of the compaction to hold the weight and do it as efficient as possible, that 1g bridge was light and efficient but breaks the spirit of the event. For that reason I understand the rule change, this is a better alternative to the second tier rule for premature braking. This ruling makes teams come up with designs that will have to be strong and hold a majority the max weight if they want to win. With the current rule in place why even have limits on how much it can hold [15kg], if the same score is produced for a bridge that weights 1g and holds 7kg and a bridge that was 10g and held 70kg under current rules both are the same.

Besides the rule for second tiering all the teams that did not hold the max 15, I just came up with this but it makes sense too without having huge score either: My score system = mass held / mass of structure then from this score subtract .1 points for every gram the structure fails to hold.

ex: mass of bridge = 10g
held mass = ..... 13.125kg
old score = ...... 1312.5
penalty = ........ 187.5
my new score =.. 1125


I don’t make the rules but this could have been an alternative to the rules that they may use next year

Re: Elevated Bridge B/C

Posted: May 26th, 2010, 7:34 am
by andrewwski
I won't disagree with that objective, but if that is indeed the objective, the execution is wrong. If there is a target weight, a second-order relationship still isn't present anywhere in the bridge. And it's not greatly penalizing those who break early if they are close - the difference between 15 kg squared (225 million) and 14.8 kg squared (219 million) would be offset by less than a 3% reduction in the bridge weight (in most cases, only a few tenths of a gram).

While this certainly does give some reward to those who hold the full load, it is hardly significant.

I'd rather see a system like nejanimb described, where there is a minimum load that is different from the maximum.