Truss Design

[KOTOR1]

I have a condition in a church that needs me to design a king post steel truss spanning 96' feet. The truss is going to carry a movable wall below and concrete roof tile on top. My coworkers is saying that I could use double angles as truss members. However, I want to know if any of you guys had designed a truss with such a long span. Is it possible to do the truss w/ double angles.

Would RAM Advanse be a good tool to design the truss or is there any other program that could do the truss design?

 

[hokie66]

The type members to use depends a lot on the configuration of the truss, the bending imposed between panel points, and on the amount of bracing provided normal to the truss. You might consider double angles, structural tees, wide flange sections, or closed sections. Program? I would do it by hand and check it by a program if you must. Truss design is fun.

 

[Ron]

Considering your span and potential loading with the moving wall and the concrete tile, double angles might be a stretch. At the least, they will be large double angles.

As hokie66 noted, you have a lot of options for members, depending on configurations.

As for the analysis, most FEA programs don't handle trusses very well without a lot of care in the modeling. If you do it, be careful with each member's restraint conditions or bracket the extremes, knowing you'll be somewhere in the middle. Welded trusses are more likely to act like an FEA model than other connections, but it isn't usually necessary to develop the moment in such connections. Do as hokie66 suggests...hand calcs and check by program.

 

[cap4000]

I have Ram Advanse and I have done many trusses with it. I would also recommend you get the book 2007 Steel Design by William Segui. Chapter 6 is devoted to trusses and he designs the connections as a beam-column with moment IF the point loads fall between the panel points of the truss.Try a WT section instead of large angles. Good Luck.

 

[Lion06]

I use RAM Advanse for custom trusses all the time. Take a lot of care in making sure the unbraced lengths for bending and compression are accurately input in the model. If you do nothing, it will assume all unbraced lengths are node-to-node. Also, take some time to look at the model with all of the end releases shown, then all the support conditions shown, then all of the loading shown, just to make sure it's all input correctly. I would also look at the actual sections to make sure they are oriented the way you want them (e.g. if you use WF's for the chords, you probably want them rotated 90 degrees so that the web is horizontal, not vertical like a beam).

 

[ahmedhegazi]

you can use double angle and i normally ,but if you have loading between cords(upper or lower) intersection with vertical or/and diagonals ,i would not recommend it,you can use many software,be sure you are driving the software well for restrains,buckling length and most important is end support conditions

 

[vandede427]

unless the software has a truss wizard to help you creat the model, be very careful with each members restraints and effective lengths, as the others have said.

i agree, double angles might be a stretch. i used double L8x8 and L8x6 for a 80 long truss, but it's depth was a constant 10 feet so the compression members weren't that long. with a pitched top chord, your web members are going to become quite long.

you might try using W shapes. rotate all truss members so that the strong axis is in the horizontal direction. this provides great out of plane stiffness for compression members, keeps the truss from being wobbly during erection, and allows you to use a variety of member sizes while maintaining good connections. for example: W12x65 top and bottom chords, W12x22 and W12x26 web members. members are bolted together on each side to their flanges b/c they all have relatively same depths (horizontal truss thickness) make sense?