Well, inspecting above sketch leaves you with no choice. You got to do the overall equilibrium equations first and then hope for the best.

From time to time there is also more than one path to happiness.

Equlibrium of .. | gives value of forces ... |
---|---|

Entire truss | R_{1} , R_{2y} , R_{2x} |

Joint A | F_{AB} , F_{AE} |

Joint B | F_{BC} , F_{BE} |

Joint E | F_{CE} , F_{EF} |

Joint F | F_{CF} , F_{DF} |

Joint C | F_{CD} |

In principle I am done now. But I have not used one equlibrium equation
on point C and none at point D ( = three unused equations because of
the three global equilibrium equations I used to start with ).

So, I would use these three unused equations to see whether they are satisfied
for the values of the forces I previously had obtained. If NOT >>>> I better
find my error.

In important note : Only in the rarest of circumstances will you find that
your check-up equations are exactly satisfied. There always will remain
some error due to round-off. So, your problem will be to judge whether
your errors can be explained away by round-off or are due because you
made an honest error somewhere.

My personal strategy of judgement is the following : If I calculate
all forces (and angles etc.) to 3 digits accuracy I can at best expect
to have my check-up equations satisfied to about 0.1 % (0.01% for 4 digits
and so on). My relative error for each equation is calculated the following
way : if I get for example 0.5 Newtons (instead of 0 ) and my largest force
is 200 newtons my relative error is 0.5/200*100% = 0.25% . It is highly
probable that this error is due to round-off. If I would have gotten 2.5 %
I would feel very uneasy and check my results.

Last revised: Aug.5 , 1997

Zig Herzog