I need to get to sleep, but I got busy with actual work, and I am trying to catch up. I think there are other issues that complicate the conversion of an orthogonal system to a radial one, but they are solvable; though I expect there will be a premium to be paid for this. Never discount the impact of "pain in the rump" factor on price. I am pretty sure we could get a contractor to play ball and give us a quote if we put together enough sketches.

120' is a big roundhouse. I could see IRM needing it this big, but how many groups have more than three engines that need this much space? The rear of the building will need more structure, and a building this big has to have a clerestory or roof monitor for lighting given how large the roof will be. Unfortunately, the additional structure at the rear and the moment frames at the inside wall need to be built to deal with the funny angles we have between bays. This means deviating from the system the manufacturers are equipped to produce. Like I keep saying, it can be done, but I would like to pick their brains for suggestions since they are the ones that would have to actually make the pieces.

I have a plan here somewhere for a 120' roundhouse. I'll dig it out and give it a look.

CH1 definitely will be replaced. It does not provide any security, and protection is deficient in many ways. A particularly disgusting problem is that pigeons roost in the roof trusses and deposit a layer of guano on the car roofs.

We are starting to discuss three major points:

1. The design and purpose of the new building.

2. What do we do with CH1's contents during construction (probably build another storage building (CH4) first)?

3. What about museum operations during construction? The current boarding area will be gone, and a lot of wire will have to come down. I don't know if we will even have trolley power into Car House 3 since its ladder tracks are very close to the job site.

I think tracks to the Visitor Center are almost a certainty.

If I might offer an idea...

A new building could be constructed between CH1 and the visitor center. This building would be entered from the middle, directly across from the visitor center, and would be able to hold everything in CH1. Once done, CH1 could be taken down, and an extension could be added to the new building on the site of CH1. Upon completion, the cars could be spread out within the building to create enough space to view the equipment, install exhibits, host events, and recreate scenes from the history of Bay Area rail transit.

If the decision is made to build a more visitor-friendly exhibit hall, I would recommend designing it from the inside out. What is the overall narrative the museum wants to tell with its collection, and how can the collection be organized and displayed to tell this story? Answering these questions should actually be the first steps in the process of formulating any plan for WRM's development.

Between the visitor center and existing CH1 at the WRM is not a location that I had considered and it does present logistical and technical problems. However the suggestion is a reminder of the need to find assistance to help us think outside of whatever boxes we have constructed for ourselves. I fully expect that what we finally do will be something I haven't imagined yet.

Back to the roundhouse, if we may...



I'm not so sure it makes all that much difference, at least as far as a steel framed pre-engineered building goes. Keep in mind that the amount of taper in each bay, 6*, isn't that much... it's only 3* off centerline, giving the bay 87* and 93* angles at the corners. In my admittedly limited experience with this sort of construction, the main frames are set parallel and the roof purlins rest perpendicular on top, and are bolted to the top flange of the frames. Each bay is then divided into a series of square panels that use tension rod diagonals to hold the roof structure square. If the main frames are 6* from parallel, the purlins can still be set square to the bay centerline, and the diagonals will work just as well.



In choosing the 120, length as the design parameter, I took a page out of your playbook... that equipment needs a space larger than that in which it barely fits to be viewed effectively. I chose an AMC Berkshire to design around, Lord knows there are enough of them around. This engine is either 99 or 100 feet long over the coupler pulling faces, depending on the source. If the engine is positioned to allow a bare minimum 3' aisle behind the tender at the doors, that leaves 18' (less the thickness of the rear wall) for people to back up and get that classic wedge shot.

I also assumed that most museums will want an example of a 4-8-4 in their collection. This does beg the question, do you design for what you have, or what your acquisition policy says you should have; I say the later. My two example Northerns, the GS-4 and J are 109' and 110' long. They will fit, not as ideal for viewing, but they will fit comfortably. As far as IRM is concerned, their Santa Fe 2903 will not fit, because of the huge tender. The IRM web site states the engine is 121'-7" long.

I also thought it would be worthwhile to explore the issues related to the long roof span near the rear wall in a building of this length. It would be useful even if it were only a portion of a larger house with additional shorter stalls.

As to a monitor roof design, it may be best to look at this as three separate pre-engineered buildings; a tall gable roof section in the center, flanked by two lower shed roof buildings, each with it's own frames. Yes, this does put columns in the aisles between the tracks at appx. the third points, although the monitor may be able to be shifted toward the doors, since the windows in the rear wall will be a major source of natural light.

For that matter, two shed roof design buildings, one a 40' span near the doors, and an 80' span building with windows in the end that extends above the lower roof might be a better bet. This puts the columns 80' from the rear wall, down where they will be between the tenders, and the shorter building can be easily designed for a greater snow load, which is common where a wall rises above a roof, to account for the drift that is likely to develop.

It does not HAVE to be round, you can have long and short stalls.

-Hudson

Hudson,

Very nice plan for the redevelopment of the Jackson St. roundhouse. Groups that have a real, live, actual roundhouse to start with are way ahead of the game. However, I thought the purpose of THIS exercise was to come up with the most cost effective (read cheapest) way to turn a virgin corn field (or any other "greenfield") into a representative roundhouse for museum display use, and hopefully do it with pre-engineered building components that would make it adaptable to several organizations needs.

Although, the more I think about the modular nature of the pre-engineered buildings, the less I agree with Mr. Ruiz that offests in the rear wall will increase costs. They will, but that increase will be offset by the savings of not building unneeded square footage to house shorter engines.

OK, I'm very interested in seeing the directions this discussion has explored so far.

Maybe we need to think radially as well as circumferentially insofar as modularity and expandability is concerned. Also, just the frame and roof without sheathing can get stuff out of the direct weather until funding for walls can be raised. I'd bet that it would be easier to raise with the skeleton and roof in place too.

I'm involved with a group whose mission is shortline and industrial interpretation - we don't want anything like a 4-8-4. Light decapods were the biggest show in our area of interpretation. No pretense of being a national collection or a mainline railroading center.

So, why wouldn't the center 75% of Dennis' 4-8-4 roundhouse be good for us?

I think the limiting critical dimension is the minimum width of the center bay - must be large enough to let a locomotive in and out. The budget's the limit beyond that..........

Yes, there needs badly to be a outer area beyond the end of the stall tracks and within the walls for just hauling stuff around. I'd not go any less than 12 feet.

Sheds and lean-tos function as well today as restrooms and support spaces as they did 100 years ago. Machine shops added on by extending tracks or having a rectangular addition - or by simply not laying track into part of the circle - are all based on historical reality.

Now, it's getting more complex and custom - but adding a rectangle in preengineered modular is a pretty well known and costable product. Might not make it more expensive on a square-foot basis.

dave

Our group, the Society for the Preservation of Carter Railroad Resources, aka SPCRR or the railroad at Ardenwood is now 8 weeks away (and a fire road) from occupying our first building/car house. This has been a 28 year long process.
We have built a 3 track, 50’ x 140’ car storage building. It is not a shop. It is not a museum. It is not public space. It is a barn to house as much of our collection as possible, out of the weather.

Our site (long term lease from a mostly benevolent 2 county park district, which they in turn lease from a local city on a 99 year lease) has a high level of regulation, and California building codes, and two agencies to please.
The building was a stretch for our group. It is a $500,000 building for a group which previously did $50,000 to $75,000 car restorations. We are a volunteer group, which means project management is difficult. (in some ways I envy the knowledge base at Orange Empire and Bay Area Electric, who have a track record of multiple significant buildings…)

To get the building, we stripped it down… it is a steel commercial building… nothing pretty. It has to be built to local seismic standards (we are 5 miles from an active fault line, 15 miles from the San Andreas fault, earthquakes are a given here). Our building required fire sprinklers… a good, no great idea… but an added complication. The building was built under California’s environmental process, and was CEQA building (California Environmental Quality Assurance)

By declaring the use to be storage, we avoided restrooms, and ADA, and had to have a lesser sprinkler system. This limits how we can use it, but we have a building, and we have a site plan for a complex… 3-5 buildings eventually, This car house, a similarly sized shop/museum nearby, an arcade depot (with restrooms) minor storage buildings, maybe an engine house… (the site use plan is already CEQAed)

Lesson 1 - We had to strip down our expectations and dreams to get something, anything…

Lesson 2 – This building is not the end of the line… there will be a second and a third, and the second will not take 28 years…

Lesson 3 – Now that we have built a building, a second does not scare us as much, and we have a track record with our supporters.

Lesson 4 – Modern building codes, governmental process must be respected… we get better buildings out of the process.

Lesson 5 – We eventually hired a general contractor with steel building experience… if it had been a more complicated building we would have needed an architect… Professionals bring much to the table…

I may think of other things as this thread continues...

Randy

Pretty much, but I'd still want to see what the manufacturers say. Remember that at the inside wall, we will not be able to brace any of the openings since they are all doorways. There will need to be another set of frames along this side to provide lateral bracing. While I am not concerned about the 3º crank for the purlins, anything connecting larger structural members is going to need some attention.


CSRM's roundhouse dispenses with economy and uses really wide angles between their tracks for this reason.

For the 120' roundhouse we are now talking about, it will need monitor or clerestory windows to light its deep interior. I have made a sketch based on Southern Pacific's 1929 Taylor Yard roundhouse, which maintained a 120' clear interior. A larger addition was planned to house the ACs, but was never built. Since this was in Los Angeles, they didn't bother with doors, and larger locomotives were allowed to hang out the back. You may notice that there is an extra girder spanning between the frames 17'-6" from the rear of the building. These support a beam between each frame. This keeps our purlin spans to 25' or less.

In this design, the bulk of the roundhouse is a shed structure that has a lean-to toward the inside to create the clerestory. I feel it is reasonably efficient. The frame could be done in one shot without the interior columns, but the extra moment connections just add cost, and roundhouses typically had a lot of interior columns anyway. The intrusion of these columns at the clerestory, and back near the tenders is not that bad.


About that wall jog issue, I still think it is a bit more complicated than you think, especially with this type of structure. The wall girts need a piece of vertical structure to attach to, so there will need to be additional columns added at inside corners. Keeping everything the same may make the labor cheaper. As I suggested, contractors charge for pain-in-the-butt costs. I could be wrong though, and it really depends upon how much building area we are talking about saving. At some point, it will make sense. Typically, a building system approach works most efficiently when it is repetitive.

In any case, designing for the deepest stall first is probably necessary if you want the roof to be continuous. If you design the small stalls first, you may end up with the roof coming down too low unless you plan for it from the start.

Thank you very much for giving some form and substance to my thoughts. Hate to admit it, but these posts have been pretty much a "stream of consciousness" affair, and after I addressed your monitor design, it occurred to me that the clerestory design has some advantages, as you've shown. Indeed, my last paragraph, which you've interpreted here, was a last minute edit on my part.


These buildings seem to be braced with tension members because steel rods are a heck of a lot cheaper than compression members. The main cross bracing for the front of this building is going to have to be at the front wall of the clerestory. At 30' high and just under 18' wide at this point, the tension rods can be at almost 45* and not intrude on the clearance diagram. Adapting the rod terminations for the 3* wedge shape would not seem to be too much of a big deal. A secondary advantage of tension rods is they can run past windows with no effect on the light, so allow the max. glass area without any custom designed sash. Hopefully, L section corner braces in the door openings will be enough to satisfy the structural guy. These will need to be custom, I'm sure


My thoughts exactly.


In my experience these buildings do not use a clear span frame in the exterior wall; formed section columns on about 15' centers is cheaper. The way I envisioned the jog detail is to put the first shorter frame on the wall line of the jog, and the rear leg of that frame becomes your corner post. The exterior wall of the jog is then a post and beam extension on the same line as the short frame. Hopefully, column length is easy to customize so that the top of the frame matches the frames of the longer stalls. If not, it may be more cost effective to make up the difference between standard lengths with the height of the foundation. More than one way to skin a cat.


Again, exactly my thinking.

A brief discussion about roofs. I know Mr. Ruiz is well aware of this, but for those of you who have been following along:

The absolute cheapest wall and roof covering is the pre-finished formed steel panels that we see on everything these days. It's nailed to wood frame agricultural buildings, screwed to these pre-engineered steel buildings, and is even available in a deep reveal style to hide the fasteners for more upscale walls. The reason it's used on everything is its low cost... the only thing cheaper is fabric.

This material is not going to work on our roundhouse roof, because it relies on the overlapping ribs to shed water from the joints. There is no way to sheath wedge shaped roof sections and seal the joints where the ribs run into the edge of the section. So, as I see it, the major increase of cost of a roundhouse vs. a plain jane rectangular building is not the framing, but having to upgrade the roof system to something like a rubber roof, and then trim all the sheathing panels to fit the odd shape. The deck can still be steel, likely one of the box section galvanized products designed to span a long distance and still present a flat surface to the roof membrane, and it is going to need a layer of composition insulation to protect the membrane from the sharp edges. This isn't all bad; a rubber roof is less likely to leak and the insulation will make it less likely for the roof to sweat and drip on the equipment, but it is going to add to the cost.

By the way, Randy, what engine did you use for an illustration, and what is its overall length?

Dennis, the engine is a Canadian Pacific 4-6-4--specifically, a class H-1b. This series is where the restored 2816 is from.

http://trainiax.net/drawings/steam6/mlw4-6-4cph1bl.gif

Tracked down the source:

http://trainiax.net/medrawhist.php

http://trainiax.net/mescaleloco.php

Other data:

http://www.steamlocomotive.com/hudson/?page=cp

She's 91 feet long with her tender, based on the entry below:

http://en.wikipedia.org/wiki/Canadian_Pacific_2816

We know her now in Canadian Pacific "Wine," but she looked good in black, too:

http://en.wikipedia.org/wiki/File:Jss1.jpg

Have fun.

I confess I had not thought about using the tension braces above the track at the clerestory. Interesting idea, and the roof bracing might be capable of transferring the lateral loads from the front wall back to that line, but then again, it might likely need its own set of frames. At a certain point, these ideas begin to become academic. Sketching out a rough outline and trying to think through the issues of a building like this is as far as I feel comfortable going without talking to some manufacturers.



Yes, of course, but it is still a complication. Simple = cheap. On this big roundhouse, the back of each bay would need two of these vertical members to frame the windows, hold the girts for the siding, and allow for back doors and/or oopsies. We would not want to have one of these verticals centered in the bay.



Pre-formed metal roofing can't be completely ruled out. Radial standing seam roofs exist. Many domed roofs feature standing seam metal roofs. That said, a membrane roof would be a fine choice, and the approach you suggest would allow for the use of rigid roof insulation, which is a better alternative than the manner in which pre-engineered buildings are usually insulated. As for the look of the membrane, roundhouses often used "built-up" asphaltic roofing, which is not known for its great appearance. Truth is, these roofs are not that visible from the ground.

Still, I'd like to hear what the manufacturers come up with.


Um, don't take the image too seriously. I eyeballed it, and it is not to scale. I was not sure how long the Hudson was supposed to be, so I made it 85'.

Thanks. I suspected it was Canadian, but I'm not that familiar with their steam.



No Fair! No wonder it looks like it rattles in the building... Kinda like the real estate pix that are always taken with a wide angle lens to make everything more spacious.

At 91' long, it's really short for a modern mainline engine; it carries a short tender. Part of our display space problem is the huge tenders that became common towards the end of steam, at least on roads that did capital improvements on facilities to allow their use. Sure hope no one finds a PRR T1 hidden in a barn, at 122'-9 3/4", it won't fit any better than ATSF 2903.

This conversation reveals to me how little I know about steam. I am unfamiliar with that term you used... "unneeded square footage" :)

Though if you are trying to build a standard building, yet some clients do not want to pay for 120' lengths, i would say, cantilevers are your friend. Set the vertical structure for a 100'ish building and tough it up enough to bear another 10' of cantilevered building out beyond it. Omitting the cantilevers will not require serious engineering, eh?