help with figuring combinations 5

[bnmorgan]

Ok, first off, I'm not an engineer, so that may be part of the problem.
What I'm trying to figure out is how many combinations of parts we could have.
We have 4 parts. Those parts are assembled togather, and only one of each part is used in the assembly. The parts are made in an 8 cavity mold. I need to know how to figure out how many combinations there possibly are. example:
part a1 a2 a3 a4 a5 a6 a7 a8
part b1 b2 ...
part c1 c2 ...
part d1 d2 ...

so i could have a1 b4 c7 d1

etc. like i said, only one of each a b c d part. If somebody could explain how to do it on paper or in excel or any which way within reason, I'll be most grateful. Thanks!!

SolidWorks 2007
NX4
Mayline 1947

 

[rbmaday]

I believe that it is the number of cavities raised to the power equaling the number of parts, or:

nc^np

 

[Timelord]

Simple! for an assembly, you have 8 choices for the first part, second part 8 choices, third part 8 choices, and fourth part 8 choices means the total combinations = 8 x 8 x 8 x 8 = 4096

Timelord

 

[bnmorgan]

thank y'all, 8^4 gives a reasonable answer. and the whole reason is they want us to test every possible combination of parts from all cavities to ensure compatability, and we're trying to use this figure to tell them "no way". each test takes ~1day, with 1 person seeing to it full time. and the test is so insanely mind numbing, nobody would survive it. we've already got a robot doing the most repetitive part. but that's all beside the point.

Thanks again!!

Byron Morgan - Tupelo, Mississippi
SolidWorks 2007
NX4
Mayline 1947

 

[SteveWag]

I think it's 1680 units (complete). He said that no two of the same part will be used in any assembly. That makes 8 x 7 x 6 x 5 = 1689.

Steve Wagner

 

[rb1957]

don't ... the robot won't survive either ... it'll go insane and bring havoc unto those who mistreated it so ... get an intern (they are expendable, and tend to wreck less havoc when they run amok).

 

[SteveWag]

Wait, I mis-read it is 8 to the 4th power, 4096.

Sorry

Steve

 

[GregLocock]

An alternative to testing every combination would be to perform a Monte Carlo, or a designed experiment (Taguchi).

Either should give you the approximate statistical data you are probably after.





Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.