Most 'efficient' truss type for a pedestrian bridge 2

[bugbus]

Hey folks,

I originally considered putting this topic in the Bridges forum, but I figured there's probably just as much experience with trusses in the main Structural forum.

Recently I have been working on some concept designs for steel truss pedestrian bridges with spans in the order of 20-40m (65~130 ft). These will be fabricated from square or rectangular SSHS with all the joints welded (typical T-, N- and K-joints). The top chords would be laterally unsupported (i.e. these would be half-through trusses with no sway bracing).

A couple of additional particulars: decking typically FRP mesh panels; clear width of walkway around 2-3m (6~10 ft); live load is 4 kPa (~80 psf).


I was wondering if anyone has any insights as to the most 'efficient' form of such a truss?

Specifically I'm referring to factors like:
*Minimising overall quantity/weight of SSHS required
*Minimising fabrication costs
*Most effective span/depth ratio
*Constant depth vs variable depth (is there a certain span beyond which it would be more economical to consider curving the top chord)?

A few articles on the internet I found seem to narrow it down to either the Warren or Pratt truss (below for reference). My first instinct is that overall the cost and material saving of one vs the other would be pretty minor in the grand scheme of things. But would be great to hear some of your experiences.

Thanks in advance for the responses

 

[GregLocock]

Neither of those is particularly efficient in terms of material usage for uniform loading. Parabolic/Bowstring is. A Cremona diagram reveals all.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[bugbus]

Greg, I suspect you're right about the parabolic/bowstring being most efficient. Especially for longer spans, I think it would be worth the additional fabrication cost of rolling the top chords, and cutting all the web members to size.

However for shorter spans, I think there's an argument for keeping it simple - constant depth, all members basically the same size, repetition in weld details etc.

 

[JStephen]

https://www.conteches.com/Bridges-Structures/Truss-And-Girders/Continental-Pedestrian-Bridge

I see a link to a "design center" that might be of interest.
Minimizing fabrication labor cost would be one of the issues, but that may vary considerably depending on who fabs it, so you may not find a "best" option until you choose a fabricator as well.

 

[1503-44]

Maximize the tension members in numbers and length
Keep compression members few and short

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[SlideRuleEra]

Pratt or the slight variation Flat Pratt (vertical end posts to easily accommodate full span handrail).

Read why here:

The Pratt Truss Explained

and

Truss - Flat Pratt

... largely for the reasons 1503-44 gave.

Also, look at the pictures of pedestrian bridges on the Contech page and other websites. The vast majority of the type bridge you described are Pratt or Flat Pratt.

 

[ClearCalcs]

When I was designing aluminum footbridges, we almost always went with Warren pony trusses. This was the most efficient structurally as diagonals in welded aluminum trusses are typically governed by tension members due to the weakened heat-affected zone.

Keep in mind you will typically need guardrails. The chord makes a nice top rail if it's at the right height, otherwise you'll likely be attaching these to the diagonals. In the latter case, handrails may need to be stronger with Warren trusses since the span is longer. Similarly you may need more stringers with Warren due to the longer span between floor beams. But then if your decking panels are beefy enough, you might be able to skip the stringers altogether and have the decking directly on the floor beams. It really depends on the individual case.

From experience, in-chord lighting (required more and more often) looks much nicer at night with Warren trusses as there are less shadows.

Span-depth ratio is interesting for pony trusses. After a certain depth (usually 5-8 ft), your compression chord buckling capacity ends up decreasing faster than the chord loads decrease from the increased depth. So if you deepen to the point that you have the head clearance, you would typically switch to a box truss. Though at that point it's often vibration that governs anyways, so you could need to upsize your chords far beyond needed for compression capacity.

It's pretty easy to add a small camber in the shop for a bridge fabricator. You can put shims at midspan and clamp down the ends to give a small camber. Diagonals are cut to accommodate the size (CAD makes this easy). The welds lock the camber in. For larger curvatures, you would want to use a proper section bender facility. This adds $$$ for shipping to/from the bender + cost of bending itself. Typically the shipping is the more expensive part. I've heard of some trying to do this in-house with chain hoists, with... not great results.

I worked in design-build so we also directly considered fab costs. Warren minimizes the number of cuts, welds, members to move around the shop. On some bridges this was the controlling factor on total project cost.

Keep in mind that a small metal pedestrian bridge might very well go for 10-20k and competition is stiff. The game is all about optimization. Quite a few times we went with heavier sizes than needed, because it was more advantageous to use up inventory than save a few pounds.

-Laurent
PS: If you're in Canada - check out CSA S7, it's brand new and has some good guidance for these bridges

http://www.ClearCalcs.com

 

[human909]

Agreed with the above. Sure Parabolic/Bowstring trusses are more efficient, I think there are even some other weird ones that are even more clever.

But we live in a world of practicalities and for a 20-40m pedestrian bridge economical design is what drives things. Repeatability, straight chords, easy connection is what reduces cost.

From observation I've seen Warren trusses used quite often with HSS. It makes sense as HSS is great in compression, in tension your connections become more critical in comparison to open section due to the lower linear space available for weld. I've done several open section trusses recently of 25-30m range of pedestrian type loads. I found underslung Pratt to be the most efficient. I looked at Warren, but to keep angles sensible the effective lengths of the chords became too great for my purposes.

The 'old skool' way for pedestrian type trusses was providing a top chord that also functioned as a hand rail. I've generally found this as inefficient. It might have made sense 30-50 years ago when labour was cheap. But I've found it easier to have my truss members BELOW the platform. Of course this is only suitable if clearance isn't an issue.

 

[bugbus]

Thanks everyone, I appreciate the really great suggestions. I won't have time to respond to each comment individually, maybe if I get some more time at a later date, but I'm taking everything on board. This has been a great discussion so far.

 

[dik]

Is there ever a need for emergency vehicles to use the bridge? Parabolic and bowstring trusses are likely a little detail complex.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[bugbus]

dik, it seems to vary depending on the client and specific site. I’ve seen this as a requirement about half the time at least from my experience.

 

[dik]

Yup... just as a caution.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[dauwerda]

From the HSS fabrication perspective, a Vierendeel truss would probably be the simplest to fabricate (if the web members are smaller than the chord members). It would definitely depend on loads and spans if it could be practical or not though.

 

[BridgeSmith]

In the modern world of bridge design, the most efficient use of materials is rarely the most economical. Today, fabrication labor is the primary cost, and material cost is a distant second place. Simple fabrication trumps lighter/less material every day of the week. That's why real truss bridges are rare, except where the design decisions are driven by aesthetic concerns. If all you want is the most economical bridge, a steel girder or prestressed concrete girder bridge is probably the way to go.

For a truss bridge, the one with the simplest and least number of member connections will win the economy race.

 

[bones206]

I live near this bridge, and I love it. Gives me joy to look at, even when I'm just driving by. Efficiency aside, something about the gentle curve makes it so much more aesthetically pleasing compared to flat chords.

 

[CANPRO]

Today, fabrication labor is the primary cost, and material cost is a distant second place. Simple fabrication trumps lighter/less material every day of the week.

This is bang on (coming from the perspective of a bridge fabricator). Using a truss for a typical pedestrian bridge is an aesthetic choice, not structural. Consider the labor in cutting/fitting all the HSS vs using a rolled girder section or even building up a girder out of plate. You're paying a bit of a premium in the weight of the bridge, but the labor is far less. Also, the cost/lbs for plate is less than HSS. Additionally, if you require notch-tough steel or fracture critical material it becomes increasingly difficult to source in rolled sections (especially HSS). Plate is readily available (relatively) as fracture critical.

Another consideration is finishing the bridge. If it is galvanized, the truss depth (combined with the length) might make it such the piece can't be galvanized as a single assembly. Designers should be somewhat aware of the local galvanizing capabilities. Any welded connections made after galv will be less durable with the typical touch-up methods. Bolted connections might ruin the aesthetics of a truss bridge. If the truss is made of HSS you'll end up with a whole bunch of vent/drain holes that will reduce the aesthetic benefits of a truss bridge. If the bridge is painted the girder option is cheaper as well. Far more efficient to paint a few large flat surfaces vs painting narrow HSS with multiple sides.

Regarding emergency vehicles, or vehicle access in general - my governing code for this would be CSA S6-19 (or S7 as ClearClacs notes but this requirement is the same) and that code requires a pedestrian bridge with a clear width 3.0m or wider to be designed for a maintenance vehicle. The design maintenance vehicle is given in the code; the gross weight is 80 kN (18 kips)) with a max wheel load of 28 kN (6.3 kips). I suspect other codes would have a similar requirement. Not as heavy as a firetruck, but significant enough to likely govern local components such as the deck and stringers and possibly the main members.

 

[mazuzam]

The answer to the question is not that simple. This has been described in length in this publication:

https://ascelibrary.org/doi/book/10.1061/9780784484920

(

https://ascelibrary.org/doi/book/10.1061/9780784484920).

The publication contains over 1000 charts depicting virtually all design conditions.