I Beam help needed

JustGary · Apr 11, 2006

I know this is not three Pontiacs, but it is a twelve foot clear span on just two poles of roughly five inch diameter. Note that it is not really sagging in the middle, despite having well over 1000 pounds on the span portion. The back beam had a warp that ended up the wrong way when the scouts put it on, so you see it below the front beam. It didn't change much when they all got on.

Wood is really very strong and has repeatable design properties, which make it useful as an engineering product. In our case, we didn't calculate that we needed about 1700 feet of rope to complete this bridge. No other fasteners!

- Just Gary

CurlyDave · Apr 11, 2006

While I completely agree that Eddie needs an engineer to design the bridge for liability reasons, I would not dismiss either wood or steel at this point.

Derek & bugstruck:

Just as a feasibility demonstration, not a design:

Even with 2x10 or 2x12 decking, could one get significant load sharing between beams, above that provided by the decking, by blocking between the beams? In a 10' span, it would normally be blocked at the center just to prevent warping and twisting of the beams. If the beam spacing were reduced to 7" centers, and the beams were blocked at 2.5', 5', and 7.5' with sections of 2x6 (This works out to 7" centers for the beams), the situation changes significantly.

For the rear tires, there would usually be at least two beams directly under the tires, so we can take full credit for them and the combination of blocking and decking might allow at least half credit for each of the two beams to either side of these two. If we happened to get one beam directly under the center of the rear tire, depending on tire width we would either have 3 beams directly under the tires, or very, very close (the inside-to-inside spacing of the first & third beams on 7" centers would be 12-1/2"), and again we could take full credit for three beams.

The front tires are a little more difficult, but I think the worst case would be with one beam directly under the tire and two outboard beams not under it. Again, at 7" on centers, inside edge to inside edge of the two outboard beams is only 12-1/2" . With 2x10 or 2x12 decking, I think the combination of decking and blocking would still be strong enough to allow full credit for three beams. This is not as clear cut as for the rear tires, but with both decking & blocking, the beams further out are still going to make some contribution.

This is similar to Piloon's idea, but I believe it allows better load sharing, or lets say easier blocking.

This would still be less expensive than steel and probably easier to build. The fastener schedule for the blocking would have to be detailed, but I bet 4 Simpson SDS 1/4 x 3 screws (they look more like lag bolts) in each side of each block would be sufficient.

I am not certain why span tables are not be sufficient to get a reasonable concept. While they don't cover the point load case, they are properly calculated for distributed loads, and I tend to trust published tables more than individual calculations from an engineer. Anyone can have a bad day, but a published table has been checked and double checked many times. And, maximum shear is the same for equal loads, point or distributed.

Derek, can you comment?

The way I look at this, the span table shows that a 12' No.2 or better beam can span 12' on 12" centers at 150 psf live load. What this also means is that a single beam should support 150 pounds per linear foot of distributed load. By only spaning 10' instead of 12' we pick up a lot of safety factor, above that built into the span table.

If we used one beam to span 12', it would be able to support 1800 lb, distributed. Certainly it can support at least an equal weight when it only spans 10', and in reality it can support more. So, each set of 3 beams could potentially handle a minimum of 5400 lb, distributed. While I realize that point loading near the center of the span is a more stressing case, the calculator at the efunda web site shows that the deflection for a centered point load is ~ 160% of that for a distributed load for a 10' span. (Yup, I used the steel I beam calculator, but since the equations are of the same form for any material, the ratio of the results will be the same.)

So, three 2x12 wooden beams 10' long could reasonably be expected to hold a 3375 lb centered point load. There would be three beams supporting each side of the tractor.

For a 4000 lb tractor, plus a 2000 lb implement, if the weight were evenly distributed side to side, we could put the entire weight at the center of the span, 3000 lb on each side, and still be OK.

Clearly this is not going to ever happen on the bridge. If all the weight were on the front wheels, there would be no traction, and if all the weight were on the rear, there would be no steering. Any other static loading is more benign.

And no, this doesn't include impact loads and does not leave a lot of spare capacity for pedestrians. OTOH if someone is driving on this bridge fast enough to produce significant impact loads, or with any significant number of pedestrians, the failure mechanism is going to be running something or someone over.

So, if we beef up the original thought, we get a bridge which fills the bill, and is still reasonably priced.

mboulais · Apr 11, 2006

Eddie -

You need to use steel beams, not wood. I am a mechanical engineer and have done the calculations for people before. I think your span is 20 feet ? Without doing any math I would say you are going to need 16 or 18" I-beams. This link is to a good website for beam calculations (and other engineering calculations. Don't focus so closely on the beams that you don't build a strong enough deck on top of them. If you will never have two-way traffic, design your deck only a couple feet wider than the widest vehicle that needs to cross it and set the beams so they are under the wheels of most vehicles that cross the bridge to minimize stress in the deck members.

http://www.engineersedge.com/beam_bending/beam_bending2.htm

I will help if needed, just send a pm with the info and I'll crunch some numbers.

bugstruck · Apr 11, 2006

Dave,

One could no doubt engineer this from wood and that latest design you propose would not likely fail. The questions is do you really want a softwood supported bridge that, pressure treated or not, has some risk of degradation over time. Everyone is thinking about moisture as the primary source of decay. We also discussed very old wood structures of a different species than yellow pine. Hardwood I presume. Now granted there are different treatment levels for different applications of PT lumber, and you can special order anything you want. But my experience with pressure treated is that the standard above grade material is subject to insect damage about 10 or 12 years out. Ants love it once it tires. Go with a ground contact treatment and your going to extend that rather significantly I'd think. Probably in the decades range. But steel as the primary support structure, in an exposed environment like this, would last forever, relative to you and I. Yes the decking will fail in time, but that's not the primary support mechanism and is more easily replaced and monitored. If anyone wonders if wood can hold a tractor you only have to look to thousands of barns that do just that. Won't find a lot of say 8K plus machines parked on softwood though. And that is of course a rather protected environment for the wood.

Regarding wood and the span tables. Yes, they are using a substantial safety factor in the design. But I've seen 2 x 10's (#2 hem-fir) spanning 13' achieve total failure with one person weighing <200 lbs. including tools standing on it. That board had no external visual signs that led me to believe I shouldn't place it or it was prone to fail. We always tried to cull out suspect material and cut for blocking. But that one did, and many others got pulled when we were placing deck joist as I could hear partial failures when walking the open joints. Mostly SPF though. Noise is the sign your on a bad one and walking load with one person exposes it. You do have to walk hundreds and thousands to see it with any frequency though. Wood joist systems are engineered with enough safety factor that the bridging and decking will transfer the loads to adjacent joists as you are quite aware. That's what is intended to cover the inequities in the material. Now for steel, that has never let me down if I thought it should work. It can fail, but I've never experienced it and it's much more predictable IMO. I believe the failure with steel is most often connection related.

So I guess this old Carpenter is still a steel beam guy on this one.

Good discussion Dave.

drm · Apr 11, 2006

Dave,

I do not mean to step on any toes I am just sharing my experience and thoughts and worst of all opinions.

( “…Even with 2x10 or 2x12 decking, could one get significant load sharing between beams, above that provided by the decking, by blocking between the beams?” )

Blocking will help in the load sharing but I can not remember if AASHTO design allows for this. I believe that they have a minimum requirement for blocking and additional blocking is not considered for load sharing as you describe. It may help but that does not mean the code will reward you for it. None the less I would do it.

( “In a 10' span, it would normally be blocked at the center just to prevent warping and twisting of the beams.” )

Yes.

( “If the beam spacing were reduced to 7" centers, and the beams were blocked at 2.5', 5', and 7.5' with sections of 2x6 (This works out to 7" centers for the beams), the situation changes significantly. )

Decreasing the spacing or ganging the joists under the tire location will help to support the loads. If the bridge is only nominally wider than the tractor and only one size of vehicle (i.e. this Tractor only not truck/auto traffic) uses the bridge then ganging or decreased spacing under the drive wheels may be used. Otherwise do this across the entire width.

( “For the rear tires, there would usually be at least two beams directly under the tires, so we can take full credit for them and the combination of blocking and decking might allow at least half credit for each of the two beams to either side of these two. If we happened to get one beam directly under the center of the rear tire, depending on tire width we would either have 3 beams directly under the tires, or very, very close (the inside-to-inside spacing of the first & third beams on 7" centers would be 12-1/2"), and again we could take full credit for three beams.” )

Yes depending on the width of the tires. AASHTO is very particular about load distribution and dictates the amount of distribution allowed based on the material and the beam (in our case joist) spacing and decking material.

( “The front tires are a little more difficult…”

“I am not certain why span tables are not be sufficient to get a reasonable concept. While they don't cover the point load case, they are properly calculated for distributed loads, and I tend to trust published tables more than individual calculations from an engineer. Anyone can have a bad day, but a published table has been checked and double checked many times.” )

The difference between a good engineer and one with experience they can tell by looking at the design if it makes sense. Experience is the only way you can get this. Tables are usually correct but even they can have errors. Addenda are commonly published for the design codes due to typos, inaccuracies, etc. I have a load book of load tables that every load table was typeset with the wrong axes on the charts. Opps!

( “And, maximum shear is the same for equal loads, point or distributed.” )

For the case of the point load at the center of the span vs. uniformly distributed load of the same overall magnitude you a correct. For the design of the wood beams shear from a point load (rear tire) at the end (actually d/2 from end) of the beam and the front tires towards the other end of the beam span is the critical shear case. This will produce a higher shear that an equivalent overall uniform load.

For example a 10’ span with 4 kips at the end and 2 kips at center of the span will have a shear of 5 kips at one end. The total load is 6k so a uniformly load distributed with this total load would only yield 3k at each end.

( “The way I look at this, the span table shows … So, three 2x12 wooden beams 10' long could reasonably be expected to hold a 3375 lb centered point load. There would be three beams supporting each side of the tractor.” )

The difference is moments from uniform to point loads:
Uniform load moment = wl*2/8 (w=uniform load. l=span length)
Point load moment = Pl/4 (P= point load at center of span, l=span length)

For l=10’ span and w=1klf load: M=12.5 ft kip
For P=10 kip point load (10’ span x 1 klf load): M=25 ft kip

( “For a 4000 lb tractor, plus a 2000 lb implement, if the weight were evenly distributed side to side, we could put the entire weight at the center of the span, 3000 lb on each side, and still be OK”. )

See load examples above for moving loads with moment and shear point load comparisons.

( Clearly this is not going to ever happen on the bridge. If all the weight were on the front wheels, there would be no traction, and if all the weight were on the rear, there would be no steering. Any other static loading is more benign.

And no, this doesn't include impact loads and does not leave a lot of spare capacity for pedestrians. OTOH if someone is driving on this bridge fast enough to produce significant impact loads, or with any significant number of pedestrians, the failure mechanism is going to be running something or someone over. )

As far as impact loads consider this: My tractor only travels at 14-15 mph tops but it bounces a lot at that speed. Consider when entering and exiting the bridge any variation in slope or grade will start the tractor bouncing. It will continue to bounce on the bridge. A wood beam is very flexible and will bounce with the tractor (to some extent.) Hence impact loads are transmitted to the beams.

( So, if we beef up the original thought, we get a bridge which fills the bill, and is still reasonably priced. )

Wood can be designed, just a matter of what size, how much, etc.

…Derek

EddieWalker · Apr 11, 2006

( Eddie -

You need to use steel beams, not wood. I am a mechanical engineer and have done the calculations for people before. I think your span is 20 feet ? Without doing any math I would say you are going to need 16 or 18" I-beams. )

The total distance of the bridge will be 20 feet long. There will be a support in the middle and a ledge on each side.

The spillway tables and calculations I've see suggest that 8 foot wide and 18 inches tall will handle my overflow needs. I'm doubleing this, and will have two overflows, side by side with aproximately 9 feet of width and 2 feet of height.

If this doesn't handle it, the dam is history anyway, because we've gone way past heavy flooding and I expect total failure at that point anyway.

As mentioned in a few posts, the footings will have to be strong enough to handle the load, plus extreme hydraulic preasure from millions of gallons of water along with movement of the bridge itself.

My thinking is to make the center support twelve inches thick and the full width of the bridge, plus another foot for the post of the railing. Ten feet wide by 1 foot thick and two feet above grade. The footings will be a foot deep and have three 12 inche holes drilled into the ground three feet deep.

Then ends will be similar in design, but with a 9 inch shelf to rest the beams on.

I will pour cement on the inside of the dam two feet past the opening to funnel the water in and a foot below grade.

There's going to be allot of rebar and cemtent in it so it lasts through a lifetime of storms. When I start construction on it, I'll post all sorts of pictures with the hope you gusy can spot a flaw or weekness and let me know of it while it's easy to change.

The discussion between wood and steel has been extremily infromative and interesting. Thanks. The biggest problem I have is my total ignorance on the topic, which doesn't allow me to even ask an inteligent question. I have no idea what questions I should be asking. /forums/images/graemlins/frown.gif

Thanks to you guys and your volunteering such useful information, I'm beginning to understand the complexities to bridge design and the various methods to accomplish this.

Eddie

drm · Apr 11, 2006

Chris,

My preference is to use steel over wood and concrete over both. I would build cast in place concrete buttress with end walls on each side of the spillway and place one or two precast double tees (8 or 10 feet wide each) over the span. Use “pretopped double tees with a 4” flange and if required for the point loads (I doubt it) pour an additional topping for punching shear. Add handrails and done, forever! But I tend to like concrete (precast for this kind of job) more than most. Have designed, detailed and trouble shot precast for some time and it just seams better for this type of application.

Precast would be permanent with no maintenance, painting, treating, etc. I have designed many “bridges” for homeowners over creeks and stream with spans up to 40 feet. These must be designed for fire truck (heavy highway truck) loading.

The pier in the middle of the spillway is, to me, asking for trouble from water undermining the foundations, debris hitting it, etc. I don’t like to put anything in the water flow path if not required. But that is my preference.

…Derek

hilld · Apr 11, 2006

Eddie,

why not cut the span down to 7 feet or so by using 2 supports rather than one. Yes, you might loose a foot or so of total spillway capability by placing an additional footer, but you are substantially reducing the distances your beam(s) will have to span. If smaller beams are allowed, there could be additional cost savings in that scenario as well, maybe you will sleep better knowing that the 7 foot span will support the live loads better than a 10 foot span.

Just a thought.

Derek

----------------------------- initial design
x...................x..................x

----------------------------- with 2 supports
x............x............x...........x

bugstruck · Apr 11, 2006

Derek,

Hadn't thought of that and I've set them before. Precast quad tees anyhow. Same idea, different application. Been so long now. The ones we set had steel connection plates and bearing plates cast in for the bearing points and side to side connections (welded). I was doing a 660' mini warehouse storage building about 30 odd feet deep. They were set on one story reinforced CMU foundation with reinforced bond beam cap block with steel bearing/weld plates, vertical in-cell reinforcement bar, and concrete filled cells. There was a center steel beam below midspan. Pre-engineered metal building sat on quad tee for the upper level. That was the entire deck system and it worked great. Very fast and strong. How economical is that these days? That's likely the best solution put forth, but I wonder the costs with the buttress/end walls. There is already some of that concrete costs associated with the steel support though.

How thick would the cross section on the double tees be for that span? I've seen them in bridge application and I recall them much deeper than the quads I referenced above. Feet deep as I recall.

I have a good failure story or two on some jobs I gave some field advice on, once they fell into crisis mode. I'll broach that another time and place. One involved the 440' sister building to that mini wharehouse I built. The precast in that design actually saved them from a total foundation wall failure. Still moved 20" mid length at the top of the wall and took a fair amount of the precast and metal with it. Both the pro and journeyman parties worked very hard to make that failure happen /forums/images/graemlins/confused.gif What a series of dumb events you must see in your line.

drm · Apr 11, 2006

Chris/Eddie,

I am sorry but it has been 10+ years since I have consulted for precasters. The costs are very regional depending on plant, competition and location. I would believe that a shallow 20"-24" double tee would work. I do not have my design info available. (I consult from home and try not to go into the office much; the commute is just shy of 1 hr.) If I am near I will stop and look-up the info. The design programs we used were written in house and ran off DOS (was in the business a while.)

All the connection hardware is cast in the tees. The buttress walls would be required for supporting wood, steel, or concrete. For concrete I would make them 12" thick. Set plates in the walls, place Tees, weld and done.

When I am being pressed on a site about who is at fault I simply say: Everyone screwed up; otherwise I would not be here. It takes a lot of mistakes, things getting overlooked, calculated wrong to get to the point of a failure. All the safety factors that have been referenced are there for a reason. That is why we have such safe buildings.

…Derek