Advice wanted: Technic frame for 40-foot Star Destroyer

[Kdapt-Preacher]

Hi! So, as discussed in the last couple of pages of this thread over on the Star Wars subforum, I'm thinking about how to build a model of the Super Star Destroyer Executor to scale with the official UCS Imperial Star Destroyer. This model would be about 40 feet long, 12 feet wide, and 4 feet tall. I'm working on this in Stud.io, with the understanding that the odds of anyone ever trying to build the thing in real life are effectively zero, but it's important to me that it be sturdy enough that it *could* be built even though in practice it isn't going to be (I measured my garage and it actually would fit, but I think some of my family members might have problems with me spending probably upwards of a million dollars on LEGO). My plan is for it to be basically a hollow shell of panels supported by a Technic frame, and I'm hoping you guys can give me some advice with that.

So basically the question is: The spine of the ship will be essentially a forty-foot-long Technic bridge supporting the cityscape on the top of the ship. It needs to be able to hold several hundred pounds without significant sagging, but that weight will be distributed across most of the length of it rather than concentrated at any single point. The main bridge beam can be supported by as many columns as necessary, but ideally not so many that they spoil the look of the ship (I'd like to have at least six feet or so between columns, and more would be nicer). The beam can be anything up to about three feet thick if necessary (in both dimensions--a solid 3x3x40 foot block of bricks is an option on the table here), but of course if it's that huge it'd be very difficult to work with. Price is effectively no object (anybody seriously considering building this won't care about a few thousand more or less dollars given the scale of the model). How would you go about building such a thing? What does an efficient girder design look like at that scale? Approximately how much weight would you trust a structure like this to carry (i.e., would I need more like 10 of those beams, or more like 100)? Are there any resources I can look at for designing this structure (short of going back to college to get an actual engineering degree)?

Thanks!

Edited February 2, 2022 by Kdapt-Preacher

[Leonardo da Bricki]

I would say that you should take a look at some of the massive tower cranes that other builders have made to get an idea of load bearing truss structures that are possible with Lego.

The most difficult part of this build will be more about the structure out to the sides needing proper support from the center beam.

Finally, the beam support structure you linked is not a great design. It won't support hardly any weight with a few plates for support, especially not on the scale you are looking at. I am not aware of any specific resource for truss structures, but you can try looking for stuff on ReBrickAble.

[Jurss]

If I would try, I would make main structure mostly of frames. Then braced with beams and even 11x3 panels. Also would need to think, how to add somewhere some diagonal beams.

[Gray Gear]

I think designing a model this massive digital only isn't good. Creating a 12 meter long technic frame without any real life prototyping will lead to some unexpected weakspots. When designing technic digital it is pretty likely  something will not work irl.

If you are an engineer who can predict all the problems the weight will cause using calculations then maybe it isn't impossible.

[1gor]

If I could help you a little bit, I'll be glad to; my MOCs are based on frames to enable motors and battery boxes installation, so I had noticed that frames 5 x 7, 5 x 11 and 7 x 11 are stiff, but frame 11 x 15 needs extra reinforcement, so I use it only when I can reinforce it with frames and / or beams when I can sacrifice space. Perhaps upcoming Liebherr crawler crane (this year's flagship) could give us some idea how to make big and stiff construction...

[2GodBDGlory]

My thought is that even the biggest trusses used in Lego cranes won't be big enough for this, so perhaps you'll need two or more "levels" of trusses. Basically, I'd try building a bunch of big conventional trusses with good longitudinal stiffness, and then pretend that those are normal liftarms and build a larger model of the same truss using little trusses as the component parts. Maybe connect the smaller trusses using 60T turntables?

[pleegwat]

Don't overbuild your horizontal supports, since it will make them too heavy.

Consider arches.

[allanp]

Many large builds like this have an internal box made of Duplo to reduce cost and piece count, but that's only good for large terrain where the whole base is supported by the floor or table. You could use the large bases from the latest Lego mindstorms set to build large "I" beams (ok, this font isn't great but I'm trying to say *eye* beams!) and then do a bunch of diagonal bracing. Imagine standing in the middle of the six foot span, would it hold your weight? Might need a few I beams built from a sandwich of many mindstorms base plates braced together so the can't come apart. Then connect with a bunch of triangles. Triangles are the strongest shape. But perhaps make the triangular cross bracing beams from technic beams and frames sandwiched between 5x11 technic panels. The panels are stronger than the 5x11 or 5x7 frames. But the frames could definitely be used in combination with the panels to make very strong beams. Then you could also brace these big beams with a triangular latticework of regular beams to prevent them bending. So big *eye* beams, braced with big triangular beams, braced with regular beams. Or you could just do what they did with James Mays Lego house and build around a wooden frame.

Edited January 30, 2022 by allanp

[anyUser]

You could also ask in the train section: Some people have been building train bridges of several meters length for 15 .. 20 years. I think not all of them were of the suspension type.

But, since the interior won't be seen I would probably use an iron girder to cover that length.

Edited January 30, 2022 by anyUser

[dr_spock]

I'd second arches.

[2GodBDGlory]

49 minutes ago, allanp said:

You could use the large bases from the latest Lego mindstorms set to build large "I" beams

I think the cost of the build just went from one million to two! Those bases cost an arm and a leg!

[howitzer]

What will this thing stand on? Should it also be made of Lego? Should it be separate or could it be an integral part of the build? I think that would affect a lot the design.

[Davidz90]

Very interesting challenge. Some time ago, I've built a Lego electric guitar and it used H-beam for neck. The forces were nowhere near your example, but I had a one metre beam holding 3 kg of string tension. The cental part of H-beam was 2 layers of plates, 4 studs wide. Sides were technic bricks. Overall, the best way to maximize rigidity is to minimize the number of connections; H-beam with central part made of largest possible plates, oriented vertically, seems like a good idea.

Another thing that comes to mind are studless technic beams connected by attaching plates to them; this slightly stresses the parts, but creates extremely rigid structure. 

[Kdapt-Preacher]

Thanks everyone for your advice--lots of useful points here!

 

16 hours ago, Leonardo da Bricki said:

I would say that you should take a look at some of the massive tower cranes that other builders have made to get an idea of load bearing truss structures that are possible with Lego.

The most difficult part of this build will be more about the structure out to the sides needing proper support from the center beam.

This guy's 10-meter crane was one of the things that convinced me that this was possible. I think that's an even harder use case than I'm envisioning, since that's an unsupported truss spanning almost the entire length of what the Executor would be, and it has to be able to lift itself as well. I'll keep looking for other large examples. My plan is for the parts of the frame extending out to the sides to be built along the same lines as the main spine (whatever that design ends up being). There'll be more columns off to the sides at the widest parts, so it shouldn't have to extend unsupported very far. The upside of the size of the ship is that there's tons of internal room to hide the supports; even at the narrowest parts of the extreme edge of the ship the internal space is still around 30 studs tall, so the frame can be quite sturdy all the way out to the ends.

14 hours ago, Jurss said:

If I would try, I would make main structure mostly of frames. Then braced with beams and even 11x3 panels. Also would need to think, how to add somewhere some diagonal beams.

12 hours ago, 1gor said:

If I could help you a little bit, I'll be glad to; my MOCs are based on frames to enable motors and battery boxes installation, so I had noticed that frames 5 x 7, 5 x 11 and 7 x 11 are stiff, but frame 11 x 15 needs extra reinforcement, so I use it only when I can reinforce it with frames and / or beams when I can sacrifice space. Perhaps upcoming Liebherr crawler crane (this year's flagship) could give us some idea how to make big and stiff construction...

Right now I'm thinking that a combination of frames and panels will probably be the most workable solution. Good to know about the 11x15 frames--I don't think those are available in large enough quantities to be the main components of the build anyway, but I'll avoid using them for serious structural applications.

13 hours ago, Gray Gear said:

I think designing a model this massive digital only isn't good. Creating a 12 meter long technic frame without any real life prototyping will lead to some unexpected weakspots. When designing technic digital it is pretty likely  something will not work irl.

If you are an engineer who can predict all the problems the weight will cause using calculations then maybe it isn't impossible.

I absolutely agree with you, but I don't have the pieces available to do anything about it at the moment. I have a couple hundred thousand LEGO pieces on hand, but not very many of them are Technic. This would be far easier if I could just build stuff and physically test whether it was strong enough to support a useful amount of weight. Once I have a truss design I'm fairly comfortable with I'll probably order enough parts to build a two or three meter section to test with, but even that'll be a couple hundred dollar investment at least.

7 hours ago, 2GodBDGlory said:

My thought is that even the biggest trusses used in Lego cranes won't be big enough for this, so perhaps you'll need two or more "levels" of trusses. Basically, I'd try building a bunch of big conventional trusses with good longitudinal stiffness, and then pretend that those are normal liftarms and build a larger model of the same truss using little trusses as the component parts. Maybe connect the smaller trusses using 60T turntables?

Ooooh, the turntables are a fantastic idea. I had already been thinking that 'fractal' supports might well be necessary, since at this scale even the diagonal support braces will themselves be something like a meter long, but I wasn't sure how to join them with connections stronger than one or two pins. The turntables should work very well for that.

6 hours ago, pleegwat said:

Don't overbuild your horizontal supports, since it will make them too heavy.

Consider arches.

4 hours ago, dr_spock said:

I'd second arches.

I'll see what I can do in terms of arches. That may be what I end up with by default, with multiple levels of triangular bracing at different angles tending to form that shape.

4 hours ago, allanp said:

Many large builds like this have an internal box made of Duplo to reduce cost and piece count, but that's only good for large terrain where the whole base is supported by the floor or table. You could use the large bases from the latest Lego mindstorms set to build large "I" beams (ok, this font isn't great but I'm trying to say *eye* beams!) and then do a bunch of diagonal bracing. Imagine standing in the middle of the six foot span, would it hold your weight? Might need a few I beams built from a sandwich of many mindstorms base plates braced together so the can't come apart. Then connect with a bunch of triangles. Triangles are the strongest shape. But perhaps make the triangular cross bracing beams from technic beams and frames sandwiched between 5x11 technic panels. The panels are stronger than the 5x11 or 5x7 frames. But the frames could definitely be used in combination with the panels to make very strong beams. Then you could also brace these big beams with a triangular latticework of regular beams to prevent them bending. So big *eye* beams, braced with big triangular beams, braced with regular beams. Or you could just do what they did with James Mays Lego house and build around a wooden frame.

3 hours ago, 2GodBDGlory said:

I think the cost of the build just went from one million to two! Those bases cost an arm and a leg!

Those Mindstorms base plates are an interesting suggestion. That's not a piece I was previously aware of. Unfortunately, regardless of cost, I don't think they're available in the kind of quantities this build would require. At 19 studs long it would take 84 of them end-to-end to stretch the length of the ship, and there are only about 700 of them for sale on BrickLink, so even if I bought literally all of them I still couldn't make a very thick beam (and that's without any of the side supports). But the 5x11 panels are more readily available (about 70,000 on BL), so I think those could work. With sufficient triangular bracing I think they'll be pretty strong.

4 hours ago, anyUser said:

You could also ask in the train section: Some people have been building train bridges of several meters length for 15 .. 20 years. I think not all of them were of the suspension type.

But, since the interior won't be seen I would probably use an iron girder to cover that length.

Ooh, the train section is a good idea. I hadn't thought about that, but you're right, there're probably a lot of serious bridges over there, and I think that's the closest analog to what I'm aiming for here. And yeah, I freely admit that making the frame out of steel would absolutely make more sense--it would be cheaper, stronger, and probably physically easier to build, given that I already have access to a bunch of box girders and welding gear--but it always seemed to me that using metal kinda defeats the point of building with LEGO. Given that I definitely can't afford the pieces to build this IRL anyway, I feel like I might as well go whole hog with it.

3 hours ago, howitzer said:

What will this thing stand on? Should it also be made of Lego? Should it be separate or could it be an integral part of the build? I think that would affect a lot the design.

Were I to actually build it, it would stand on a flat concrete garage floor. I was assured when this house was built that the garage foundations are strong enough to park main battle tanks, so hopefully that won't be a problem. I want everything above that to be made of LEGO. This is a starship, so the model itself will be elevated on support columns, which I intend to be maybe five or six feet tall (so the midline of the ship is roughly at eye level). The ship itself will be mostly just a thin layer of plates laying over the Technic frame, so it won't have any structure without the frame to sit on. I would love to design it so the frame could come apart in sections to transport, but if that isn't possible it's not the end of the world.

1 hour ago, Davidz90 said:

Very interesting challenge. Some time ago, I've built a Lego electric guitar and it used H-beam for neck. The forces were nowhere near your example, but I had a one metre beam holding 3 kg of string tension. The cental part of H-beam was 2 layers of plates, 4 studs wide. Sides were technic bricks. Overall, the best way to maximize rigidity is to minimize the number of connections; H-beam with central part made of largest possible plates, oriented vertically, seems like a good idea.

Another thing that comes to mind are studless technic beams connected by attaching plates to them; this slightly stresses the parts, but creates extremely rigid structure. 

This is an interesting idea as well. I'm going to have to have structures of all sizes, with the huge central spine at one end out to the little 'capillary' struts that hold the individual hull plates (which won't weigh very much, so they won't need to be particularly strong). That kind of structure would work for that. I don't have any objection to the slight stresses from attaching plates to liftarms (every piece in this build is gonna be stressed one way or another), but I worry a bit about that limiting the amount of cross-bracing I can do since it occupies so many pin holes. I'll see how it works out.

Edited January 30, 2022 by Kdapt-Preacher

[SNIPE]

The good news is you dont have to worry about weight since it is a static model. Maybe this thread could be of use:
 

Regards, Snipe

[mdemerchant]

I may be in the minority but I'm not sure building a 40 foot long beam to hold up several hundred pounds will actually be that hard if it is supported every 6 feet.  That's 45 lbs/span if several means 300.  Not that I have built that sort of structure but based on the breaking strength of liftarms and pin connections (https://eprints.usq.edu.au/20528/1/Lostroh_LegoTesting_2012.pdf) it seems like it should be very possible with the right design.  I think it's important to do some rough math and figure out how much weight you really need to hold up.  You said several hundred lbs but do you know it will be in that range or are you just guessing?

Here is a simple beam calculator that can help you to understand the forces involved in the beam if you want to try some theoretical analysis.  https://mechanicalc.com/calculators/beam-analysis/  The problem will be with 6 foot long spans you will likely have buckling and/or the structure just popping apart sideways long before you hit the max theoretical load.  You will definitely want to have it form locked transverse to the length of the beam and probably use two or three identical beams (or trusses) in parallel with cross bracing connecting them to avoid that.

[Saberwing40k]

Well, it's taken me a while, but I have finally collected my thoughts on this. This project is absolutely nuts, and I applaud you for it. Since you seem to be adamant about not using non Lego parts, I'm not going to even bother discussing a steel frame.

However, your model is huge, there is no other way to say it. It's as long as a 40' shipping container, but 50% wider, at least at its widest point. At this scale, I see absolutely no point in attempting to design a frame. There are so many potential pitfalls to designing a beam of this scale digitally, real prototyping would be mandatory. I can think of no computer or program that would be able to subject a Lego construction like this to physics testing, AND you'd need some pretty specialized knowledge to do that.

Regarding the size, part count, and weight, I looked around for something I remember seeing a long time ago.

This guy built a minifig scale model of the Japanese WW2 battleship Yamato. It is 22 feet long, 3 feet wide, and weights 330 pounds. It also contains about 200,000 bricks.

Your theoretical Executor SSD would be twice as long, 4 times as wide, and probably twice as thick. So, 16 times bigger. Now, this of course ignores the fact that the Executor tapers quite a bit, so it's not 12 feet wide the whole way down, but your model can't sit on the floor like this one can, so it balances out. From this ballparking, your model would take 3.2 million bricks, weigh 5280 pounds, and cost about $320,000, if you are able to do the average of 10 cents per part. Now, I think you'd actually not need quite that number of parts, given that the Executor would not have nearly as much detail, and large swathes of it could be made out of relatively few large panels. This is still a nice large number, though.

For some further structural inspiration, I'd look at the life size Technic builds of the Bugatti Chiron and Lamborghini Sian, which are both mobile models that are quite large, and also look at some of the stuff TJ Avery has built:http://66.39.118.91/model_17bridge/index.html

Also, you may need a bigger computer. Rendering 3 million bricks would be quite taxing.

[Kdapt-Preacher]

39 minutes ago, mdemerchant said:

I may be in the minority but I'm not sure building a 40 foot long beam to hold up several hundred pounds will actually be that hard if it is supported every 6 feet.  That's 45 lbs/span if several means 300.  Not that I have built that sort of structure but based on the breaking strength of liftarms and pin connections (https://eprints.usq.edu.au/20528/1/Lostroh_LegoTesting_2012.pdf) it seems like it should be very possible with the right design.  I think it's important to do some rough math and figure out how much weight you really need to hold up.  You said several hundred lbs but do you know it will be in that range or are you just guessing?

Here is a simple beam calculator that can help you to understand the forces involved in the beam if you want to try some theoretical analysis.  https://mechanicalc.com/calculators/beam-analysis/  The problem will be with 6 foot long spans you will likely have buckling and/or the structure just popping apart sideways long before you hit the max theoretical load.  You will definitely want to have it form locked transverse to the length of the beam and probably use two or three identical beams (or trusses) in parallel with cross bracing connecting them to avoid that.

It's a guess, but a reasonably informed one. The smooth areas of the hull will have a combined weight of about 400 to 450 lbs depending on the amount of detailing, but spread over about 200 square feet of surface area, so the weight on any individual beam in those sections will be fairly low. The dense cityscape areas on top of the hull will weigh quite a bit more (maybe twice as much?), spread over a smaller (but still quite large) surface area. I have no idea what the exact numbers will be, but I would think under 10 lbs per square foot. There may be a few specific areas with greater weights, such as the points where the engines are attached, but my goal would be that the cityscape itself be fairly rigid, such that the weight of the whole thing was supported principally by the vertical columns rather than anywhere near the middle of the spans. I would say that I'm not really worried about physically breaking pieces unless I do something *really* dumb with them, but since the total magnitudes of forces here are pretty large I'd like to play it safe, especially since my ability to physically build and test anything at this scale is highly limited. I'll play with the beam calculator and see what I can make of it.

[1963maniac]

The bricks will be expensive!! Bricks and plates still have the strongest connections. So use Technic bricks for sure.

But, you'll also need a very fast CPU, very large monitor, lots of RAM and a major GPU with it's own RAM. I wish you the best, if you decide to  continue with this BIG dream.

 

Edited January 31, 2022 by 1963maniac

[Kdapt-Preacher]

42 minutes ago, Saberwing40k said:

Well, it's taken me a while, but I have finally collected my thoughts on this. This project is absolutely nuts, and I applaud you for it. Since you seem to be adamant about not using non Lego parts, I'm not going to even bother discussing a steel frame.

However, your model is huge, there is no other way to say it. It's as long as a 40' shipping container, but 50% wider, at least at its widest point. At this scale, I see absolutely no point in attempting to design a frame. There are so many potential pitfalls to designing a beam of this scale digitally, real prototyping would be mandatory. I can think of no computer or program that would be able to subject a Lego construction like this to physics testing, AND you'd need some pretty specialized knowledge to do that.

Thank you! I know you're right, I just don't know what else to do about it. I have a couple hundred thousand bricks on hand, so I'm used to being able to immediately prototype pretty much anything I want, but I don't have remotely the right assortment to make a serious play at something like this. If I took apart all my Star Wars sets I could get together a couple hundred liftarms, but even that wouldn't get me a proper test here--I'd need to build probably a 12-foot section to really be confident that I'd tested it thoroughly. If I really get deep into this (which I'm honestly not sure I will yet; at the moment I'm still mostly doing math to try to figure out what this would even look like) I'll probably buy a few hundred dollars worth of Technic panels so I can do that, but in the meantime this is the best plan I have.

42 minutes ago, Saberwing40k said:

Your theoretical Executor SSD would be twice as long, 4 times as wide, and probably twice as thick. So, 16 times bigger. Now, this of course ignores the fact that the Executor tapers quite a bit, so it's not 12 feet wide the whole way down, but your model can't sit on the floor like this one can, so it balances out. From this ballparking, your model would take 3.2 million bricks, weigh 5280 pounds, and cost about $320,000, if you are able to do the average of 10 cents per part. Now, I think you'd actually not need quite that number of parts, given that the Executor would not have nearly as much detail, and large swathes of it could be made out of relatively few large panels. This is still a nice large number, though.

That Yamato is quite the build, and I think it's a fair comparison. My current plan is to do the entire smooth hull sections as a series of 80x80 panels made of 16x16 plates (i.e., 5x5 plates as the outer layer connected by 4x4 plates as the inner layer, for 2 plates' thickness over the majority of the surface). I've already tested that and am comfortable that those panels are light and strong enough to form the hull--they're not structural anyway, so they just have to hold themselves together and not sag much. The whole interior of the ship will be as hollow as I can make it; there doesn't have to be anything in there except whatever support struts the hull needs (probably something like a horizontal beam every 80 studs or so, to align with the hull plates). As a result, the only dense areas of the ship will be the top cityscape region and the rear surface around the engines, and I'm optimistic that it wouldn't weigh as much as your estimate. However, those areas have extremely dense detailing and will involve a crapton of tiny parts. I'll need to mock up a section of that so I can get a better estimate of how many, but there's no question that it's gonna be a big number.

42 minutes ago, Saberwing40k said:

For some further structural inspiration, I'd look at the life size Technic builds of the Bugatti Chiron and Lamborghini Sian, which are both mobile models that are quite large, and also look at some of the stuff TJ Avery has built:http://66.39.118.91/model_17bridge/index.html

Also, you may need a bigger computer. Rendering 3 million bricks would be quite taxing.

That's a useful page--thanks! Although given that that bridge did collapse I'm not sure how closely I should follow their engineering advice, LMAO. And yeah, my computer may be an issue--I already tested that and Stud.io became largely unusable after around 600,000, so this would probably have to be done in sections. I'll cross that bridge when I come to it; I'm not yet certain that I'm going to get that far.

 

 

17 minutes ago, 1963maniac said:

The bricks will be expensive!! Bricks and plates still have the strongest connections. So use Technic bricks for sure.

But, you'll also need a very fast CPU, very large monitor, lots of RAM and a major GPU with it's own RAM. I wish you the best, if you decide to  continue with this BIG dream.

It'll almost certainly be outside my budget, ultimately, but we'll have to see. It's important to me that it be possible to build, but much less important that anyone actually do so, so for practical purposes I'm content with considering this a digital art piece. But if I do finish the design, I'll definitely at least bounce the idea off LEGO and Disney... Somebody already tried and failed to Kickstart construction of a 13-foot Executor, but this one would be way cooler. My current computer definitely won't be able to handle the entire model, but my brother's in CS, has access to supercomputers, and is bribable, so I think I'll be able to work something out. Plus, honestly, by the time I finish this computer technology will probably have advanced significantly from what we have today.

Edited January 31, 2022 by Kdapt-Preacher

[howitzer]

15 minutes ago, Kdapt-Preacher said:

And yeah, my computer may be an issue--I already tested that and Stud.io became largely unusable after around 600,000, so this would probably have to be done in sections. I'll cross that bridge when I come to it; I'm not yet certain that I'm going to get that far.

There's another software, called LeoCAD which is optimized for large models with many parts, so you might want to check it out. Not sure how much further it can push the limit though, as it might very well be that your computer is just not fast enough for the amount of bricks you want. 

[rm8]

Build more triangles Out of liftarms.

I've build big structure using liftarms, and Pifagor really helps.

This is useful topic for you

 

[Gray Gear]

Wouldn't technic bricks be better suited for this than liftarms? They tend to bend a lot less

[allanp]

Those mindstorms bases are pricey but they are available on Lego bricks and pieces for £4.69

[howitzer]

1 minute ago, allanp said:

Those mindstorms bases are pricey but they are available on Lego bricks and pieces for £4.69

I think B&P limits the number you can buy though. Not sure what's the limit (might depend on the part) but I'm quite sure the limit is well below the hundreds or thousands you'd likely need for a project like this.