gantry - channel sections

Hi Jim - starting topic's own thread...
Hi Jim
Can you sketch? That is always helpful. I do that even when it seems we all agree, as it can be amazing how a sketch reveals that the impressions are not the same.
I'm thinking - maybe weld some brackets on top of the channel with the bolted flange faces facing. I'd normally wish there could also be a flange underneath - but the close-fitting "standard" carriage would collide with it, so not possible.
That said: if I understand correctly (big "if") - the loading may be "cantilever beam" - tension on top, compression underneath. Extending out of the shed? Suspended - near the door? In which case flange on top, nothing needed underneath as compresses together would be bril.
Load analysis of the section - as channels tend to twist under load, and is singular, I wouldn't dare use Euler-Bernoulli beam calculation as works for symmetrical beams.
Sorry verbal description is hard to "see".
Rich S
-----------------------------
The channels will be attached web to web, with the splice plate sandwiched between them at the center, hanger plates between them at the ends and other stiffening as needed, so they should behave as a single symmetrical I beam. So far the 3" channel track hasn't noticeably twisted under a 1/2 ton load.
Any cantilever loading will be the channel's own weight while setting up plus possibly the weight of one end post and diagonal braces if the opposite loaded side post sinks into wet ground. The model is a beam with pinned joints to columns at the ends. Assume the columns are pinned at the base and could tilt inward if the beam sags, though they won't tilt out or laterally.
Really I just need a good bolt geometry for the splice plate so I can start calculating stresses. I'm leaning toward two large bolts through the channels' neutral axes beside the joint and two more at the outer ends of the vertical splice plate, and welding reinforcements to the web around them if necessary. Perhaps a grid pattern of more smaller bolts would be better?
It seems to me that the beam splices I've seen on bridges might weaken the lower flanges if the bolts aren't sufficiently tight, and this track must be disassembled repeatedly and probably sprayed with slippery rust inhibitor wax. The splice plate I have now, from a local steel erector's scrap pile, is 10mm thick, 150mm high and 700mm long.

Can you sketch?
---------------
If you send me an email I'll attach a photo of the gantry and sawmill. it's on Google Photos but I don't know how to share it.

Right click on photo, select "Copy link address" Paste link in new message.

Right click on photo, select "Copy link address" Paste link in new message. -----------------------------------------

Jim - get an "Error 404" on trying to follow that URL. Tried immediately to get there and found that problem. Sorry didn't get back to you immediately. If U-channels are used symmetrically - back to back or facing - so that they can't twist (rotate) under loading - that a simple analysis just using 2nd moment of area, beam metal yield strength and beam length/support is useful. That also applies to "L" sections, by the way - facing or back-to-back. In a "desperate" situation I worked out the strengths according to Euler-Bernoulli beam and it seemed very right. As the welder, I made sure the secondary steelwork was less strong (would warn by taking permanent bend) and less stiff than the "primary" steelwork - so you sensed the limit - and it worked so well.
If you still would like some additional input / collaboration on this, we can try again with pix and dimensions.
I've done my best with "ASCII art" to sketch a buttressed bolted connection. That only works if the beam is cantilever, with the top in tension and the bottom in compression. The wheels of the carriage have clearance under the buttress arrangement.
_ _ / |_| \ ---------------/--|_|--\---------------
------------------- -------------------
Which I think will be stiff and strong. You could trap a "hanger plate" protruding out of the top as the means to suspend the transporter beam.
I hope I have * understood the situation * the suggestion is helpful
Best wishes, Rich Smith

> Jim Wilk>> >> Can you sketch? >> >>

---------------
I just found these to study:

I regularly use the "Blue Book" for sections. Didn't know of the "Green Books" for connections. Thanks.

I regularly use the "Blue Book" for sections. Didn't know of the "Green Books" for connections. Thanks.
------------------------------------
This is very good too:

I've returned to the gantry track splice, after fixing the car to pass inspection and largely completing my taxes.
Have you ever seen a bolted I beam splice with the flanges thickened with welded plates or such to compensate for the reduction in area at the bolt holes?

I've returned to the gantry track splice, after fixing the car to pass inspection and largely completing my taxes.
Have you ever seen a bolted I beam splice with the flanges thickened with welded plates or such to compensate for the reduction in area at the bolt holes?
--------------------------
If welding plates on the outside of the channel flanges to thicken and strengthen them at the central splice is technically (if not economically) sound, they provide much more thread engagement for splice plate bolts in tapped holes, that don't interfere with the trolley wheels rolling on the lower flange.
All the examples I've found of bolted beam splices accept the loss of net flange area at the bolt holes and instead suggest locating the splices away from the most heavily stressed parts of the beam.

A patent here that uses wedges or teeth on plates and beams that would transfer stress over a broader area:

Maybe poke around in the patents for a bit, give you some more ideas to work with...

A patent here that uses wedges or teeth on plates and beams that would transfer stress over a broader area:

Maybe poke around in the patents for a bit, give you some more ideas to work with...

... A patent here that uses wedges or teeth on plates and beams that would transfer stress over a broader area: ...
-----------------------------
I don't have the shop space to clamp one end of an 8' beam in my mill vise to accurately drill a hole pattern. All I can do is hang the opposite end at an angle in a doorway and locate a punch mark or pilot hole under the spindle to drill it larger. The splice plate will have to be match-drilled to beams on sawhorses out in the driveway. Either a Portalign or clamped-on bench drill press will keep the bit, hole and tap square to the surface. I use an adjustable hand reamer to give bolt shanks the minimum of clearance.
That's how I spliced 8' sections of 3" channel into the sawmill track and overhead gantry, but those splices have to sustain only the dead weight of the channels while I'm setting up and leveling them, they are supported at the splice in service. I'm trying to splice the 4" channel strongly enough that it doesn't need a center support although it will normally have one. The problem is designing a tension-side joint that gives the full strength of the channel without blocking the trolley.
If I just weld them together or buy a suitable 16' beam the assembly is too large to store under cover and too heavy to carry and lift into place. One 8' section of the 4" channel is hard enough to support and align with one hand while inserting a bolt with the other, on a stepladder. When possible I turn a taper and a root-diameter pilot on the ends of machine screws to help align tapped and clearance holes in things I repeatedly assemble and disassemble outdoors, like my 50' TV antenna mast, but there isn't enough clearance for that between the trolley wheels.
The problem could be worse. I had the chance to examine and ride in a Beech 18 whose original wing spars had been stiffened with a similar-sized steel reinforcing strap underneath. It makes an un-aerodynamic bulge in the lower wing surface.
"Other manufacturers kits resemble bridge trusses and can protrude more than 10 inched from the skin of the Beech. Not only is this unsightly but it is enough to make an aerodynamicist sick. I heard one owner say that one good thing about his bridge truss spar strap was that if he ever had to belly land his Beech at least the props wouldn't hit the ground!"

The problem is designing a tension-side joint that gives the full strength of the channel without blocking the trolley.
-------------------------
Am I correct in assuming that thickening the outside of the channel flange with a welded-on shim to restore the area lost to bolt holes and sizing the joining plate to at least that same net cross-sectional area is enough?

The problem is designing a tension-side joint that gives the full strength of the channel without blocking the trolley.
-------------------------
Am I correct in assuming that thickening the outside of the channel flange with a welded-on shim to restore the area lost to bolt holes and sizing the joining plate to at least that same net cross-sectional area is enough?
-------------------------------
This is similar to a filler plate in a splice between beams of different heights. Are filler plates ever welded to the smaller beam?

This is a half section of my present gantry hoist track splice design. C is the 4" channel, V the 3/8" thick central splice plate between the channels, H the 3/8" thick horizontal splice plate under the tension flanges and A the angle iron splice above the compression flanges.
My question is how to weld the inverted T joint between V and H. I think an inside fillet weld will deform H upward and require grinding off the sharp and square lower corners of the channel. Is a row of plug or slot welds in H a reasonable alternative to avoid distortion and interference?
The 6" high vertical center plate V alone has nearly the same Ixx as the two channels so I don't think welding it to H is essential but connecting them should make the joint stiffer and one-man assembly easier. The angle A on top bolts to V and C afterwards, and gives the central hanger shackle pin more bearing area than V alone would.
VA VA VA VAAAAAA VCCCCCC VC VC VC VC VC VCCCCCC HHHHHH

Is what you aim for structurally necessary or helpful is a good first question given the work involved...
OK - using your notation and adding W for weld
.. VC VC VCCCCCC WHHHHHW
So you now have the "H" as two strips with the inner edge bevelled about 30deg. Set with a root gap at least equal to the middle "V" plate. Wider and you "get" some of the "C" channel. Weld-up. Left and right corners and if necessary middle run for first layer, then fill.
If "H" is a tension "doubler", not otherwise part of the beam "Second Moment of Area", then just fillets at the ends would do? Maybe small fillets along sides, out to edges of the "C" (channel) sections?

Is what you aim for structurally necessary or helpful is a good first question given the work involved...
OK - using your notation and adding W for weld
.. VC VC VCCCCCC WHHHHHW
So you now have the "H" as two strips with the inner edge bevelled about 30deg. Set with a root gap at least equal to the middle "V" plate. Wider and you "get" some of the "C" channel. Weld-up. Left and right corners and if necessary middle run for first layer, then fill.
If "H" is a tension "doubler", not otherwise part of the beam "Second Moment of Area", then just fillets at the ends would do? Maybe small fillets along sides, out to edges of the "C" (channel) sections?
------------------- I was expecting to do just that until I cleaned up the rusty channels and found that the flanges are square to the web and a one-piece H plate fits flush on both sides.
This with an abrasive fiber pad stripped the rust remarkably fast and effectively and was easier to control than a 7" angle grinder.
The 100kg track and hoist assembly has to come apart for handling and storage so the channels can't be welded to the splice.
If the angle between H and V initially matches the outsides of the channel, would it increase or decrease after filling the gap?
If the angle opens a little the channel would still bear solidly on H at the root which may be acceptable. If the angle closes the channel will contact H at the outer edge and shift under load. I don't have a press large enough to bend 3/8" plate.
I considered plug welds assuming I could fill a 3/8" hole in one pass and the weld would shrink symmetrically as it cooled, and the surrounding solid metal would minimize distortion.
The plates and channels would be clamped together with a gap between the channels under the weld. I could stop drilling and filling if the plates began to warp out of alignment.