Small USV multihull design considerations?

Hello everyone!

I'm working on a second iteration USV platform for various academic research projects and am seriously out of my depth when it comes to picking the correct hull shape (I'm primarily a robotics guy and boat design is more of a hobby for me). I was hoping someone who knows their stuff around here could give it a peek and let me know their thoughts so I don't make any overly stupid mistakes

To outline some general specs I'm planning on, here's the general idea:
- length: 1.2m
- beam (overall): ~0.6m
- draft+freeboard: ~0.2m
- mass: 15-20 kg
- power: 1.2 kW (lithium-ion, 4xT60 thruster, differential thrust steering)
- materials: XPS structural core, glass fiber composite wrap
- speed: 5-8 kts

The previous design I built was about 0.7m and was a displacement multihull, but I seriously underestimated just how powerful BLDC motors are, and it was able to get up on plane at 5 kts regardless with only one of these thrusters. So this time I'm designing it as a planning hull from the start instead, and hopefully gain some cruise efficiency and top speed.

So far I've settled on a sort of tunnel boat shape without the tunnel, shallow vee (about 10%) seems to make sense to make it easier to get on plane and stay stable once it's there from what I've read, but I'm mostly guessing and winging it.

I've started off with the stock planning hull from Delftship and adapted it to my parameters, which ends up looking a bit like so:

I'll be hot wire cutting the XPS, so having a long straight lines is certainly best and for laying epoxy later, but something like a curve that alters the angle along the hull which I've seen a lot of tunnel hulls have would be uhh, harder.

The best idea I think I have would be to have deadrise be at a max of say 15 degrees around a third of the way from the bow which would continue straight, then steps that reduce that to 10%, 5% and 0% the stern, where I can then very easily mount the thrusters on a flat surface and avoid having to cut any curves on that part.

I'm not sure where I'd place the steps exactly or how to calculate that, would having them evenly spaced be a good enough rule of thumb? None of this has to be all that accurate, just sort of ballpark correct.

Assuming that's not all malarkey, does it make more sense to mount thrusters before (A) or after (B) the last step? I suspect the A position (before the step) would be far better with less ventilation , but I'd still have to deal with the 5% slope. In any case the area would have a lot of turbulence so it might not make much of a difference.

In general the thrusters would be mounted at some slight pitch for more lift and to better point at the center of mass, but also have a slight inwards yaw since they're mounted too far back for a good steering response with diff drive. Might add rudders at some point too or switch to azipods, but waterproof servos are their own can of worms so it's filed under maybe.

Another thing I'm wondering about is adding a bit that narrows at the stern:

It would be some of extra effort to make, and would have less buoyancy at the stern too, is it worthwhile to reduce drag a little?

I've cut out a very rough 1/2 scale version of the stepless type out of EPS to practice hot wire cutting the other day. In bathtub tests I think there's a little more drag than I was expecting (partially from all the rough cuts when trying to follow a printed template), but I can't really test if it planes well since it's just slapped together and pretty fragile.

I'll make a more proper 1/2 model out of XPS next and do some lake pull tests once I have more of an idea of what the end design should be.

Please let me know what you think. Thanks!

Hi MoffKalast, your idea of having steps of dead rise will produce drag at every step. Better to have a smooth stepless transition allowing smooth water flow back to the point where you want a flat bottom for drive mounting. Azimuth pods would be much more useful, with only one needed for each hull, reducing the drag of the blade housings. You could possibly mount your chosen thrusters on your own design azimuth steering system, using simple R/C bell cranks and servos. (not 360 deg.), to simplify things. Looks like an interesting test bed for researchers.

MoffKalast said:

...
To outline some general specs I'm planning on, here's the general idea:
- length: 1.2m
- beam (overall): ~0.6m
- draft+freeboard: ~0.2m
- mass: 15-20 kg
- power: 1.2 kW (lithium-ion, 4xT60 thruster, differential thrust steering)
- materials: XPS structural core, glass fiber composite wrap
- speed: 5-8 kts

...

Click to expand...

Just on these numbers alone, you have some issues.
The length-displacement ratio is very low, circa 2-ish....this means your resistance will be high.
Your Fn is 1.1-1.2, so given the higher drag, getting on the plane maybe problematic.

seasquirt said:

your idea of having steps of dead rise will produce drag at every step. Better to have a smooth stepless transition allowing smooth water flow back to the point where you want a flat bottom for drive mounting

Click to expand...

Ah interesting, so steps need to follow the hull shape exactly? And I thought I would be able to get away with straight sections and triangular steps... I guess I might need a better jig to cut this out as a continuous shape and plan for some sanding and fairing.

Btw, is there a way to determine what would be a good vertical delta of the flat stern relative to the sharpest deadrise? I see most hulls have the stern slightly higher up, so the center of buoyancy is further forward for more lift, but I could do anything from zero to a few cm, is there an angle I should aim for?

seasquirt said:

Azimuth pods would be much more useful, with only one needed for each hull, reducing the drag of the blade housings. You could possibly mount your chosen thrusters on your own design azimuth steering system, using simple R/C bell cranks and servos. (not 360 deg.), to simplify things.

Click to expand...

For sure, and I'm planning on exploring some more experimental propulsion designs eventually (maybe even a quad thruster ROV-style setup for omnidirectional motion) since we can just attach stuff to the aluminium extrusions, but there is added complexity in that and the reliability won't be as good until I figure out good approaches. Initially I was just planning one thruster per hull, but in the last design we've had one of them fail so this has four mainly for redundancy.

Ad Hoc said:

The length-displacement ratio is very low, circa 2-ish....this means your resistance will be high.

Click to expand...

Hmm do those estimates scale down to these small sizes well? If I calculate DLR by the wiki formula I get something ridiculously small per hull, like (10kg / 1016) / ( (1.2m * 3.28084)^3 * 0.01) = 0.0161
The mass is just a ballpark estimate at this point regardless, depends on how much payload it'll carry and how much I XPS I can replace with EPS during construction, haha.

Ad Hoc said:

Your Fn is 1.1-1.2, so given the higher drag, getting on the plane maybe problematic.

Click to expand...

Yes it seems like that by Fn, 5 kts would be transitional and 8 kts would be planing for this waterline length. I do suspect that this thing will spend a lot more time at 5kts than above (that's the max speed our current DVL can do anyway), so it might make more sense to optimize the hull for semi-displacement mode instead?

MoffKalast said:

....so it might make more sense to optimize the hull for semi-displacement mode instead?

Click to expand...

You should clearly define your SOR and design the boat first, before making such decisions.
The SOR shall dictate what type of hull form is best, and will be the last thing to consider, not the first.

Ad Hoc said:

clearly define your SOR

Click to expand...

Well, when it comes to research use there is unfortunately no such thing. Projects come and go, one day you need slow bathymetric mapping, the other day you need long range high speed, the third day you need sideways docking, the fourth day it needs to just stay still in the middle of a lake with a large board so drones can land on it.

The important bit is that as few parameters are set in stone, so it can be adapted as much as possible to the task at hand. The four connecting rods are planned to be adjustable, so the total beam is configurable, and equipment can be added, removed, or moved back and forth to adjust the CG. It's fine if it's a bit meh in most roles, as long as it gets the job done.

The stern bottom rising to just above the waterline is a way to eliminate the drag of an immersed transom, eliminating the drag of the 'last step'. The angle can be the one with the least disturbance to water flow, or a compromise with any other parameters already fixed by the design. Before computers people just made shapes that 'looked good' in their flow of the lines, so if it looks good, maybe it is.
Sea planes have one or two steps on the hull to help leave the water, but they create drag. Look at fish, slippery as - no sharp steps.

Just make your SOR list to your min's and maximum's you are likely ever to need, eg. minimum draught, 20Kg payload centred at X position(s), 8 Knots speed, 0.5m high (salt water?) waves at X frequency, stability to X standard, max hull length overall (with transportation in mind), fixed form or de-mountable, engineered grid with tapped holes for payload attachments, battery power requirements for boat and payload - physical size / shape - weight incl. placement - conductors - switches, R/C system, GPS?, mast for comms or cameras, boom arm for manipulation, will it ever need to tow or push something, etc. Go nuts with possibles, then optimise and prune.
Then you have a list of 'requirements' to compromise against each other and your $ / time budgets.
Leave room for the latest 'must have' gadgets and sensors after all your calcs. are done. Have Fun.

seasquirt said:

The stern bottom rising to just above the waterline is a way to eliminate the drag of an immersed transom, eliminating the drag of the 'last step'. The angle can be the one with the least disturbance to water flow, or a compromise with any other parameters already fixed by the design. Before computers people just made shapes that 'looked good' in their flow of the lines, so if it looks good, maybe it is.
Sea planes have one or two steps on the hull to help leave the water, but they create drag. Look at fish, slippery as - no sharp steps.

Click to expand...

Ah alright, I thought there may be some additional factor in play when it comes to planing hulls. For displacement ones the fish approach certainly seems best and looks sensible, anything from sailboats to tankers follows that principle, but I've already made that mistake with my previous design...

From what I understand the steps are mainly there to reduce wetted area when there's enough lift, and it takes less energy to detach the flow instead of smoothly curving it?

seasquirt said:

Just make your SOR list to your min's and maximum's you are likely ever to need, eg. minimum draught, 20Kg payload centred at X position(s), 8 Knots speed, 0.5m high (salt water?) waves at X frequency, stability to X standard, max hull length overall (with transportation in mind), fixed form or de-mountable, engineered grid with tapped holes for payload attachments, battery power requirements for boat and payload - physical size / shape - weight incl. placement - conductors - switches, R/C system, GPS?, mast for comms or cameras, boom arm for manipulation, will it ever need to tow or push something, etc. Go nuts with possibles, then optimise and prune.
Then you have a list of 'requirements' to compromise against each other and your $ / time budgets.

Click to expand...

Sounds like a good approach The parameters that are fixed are the length of 1.2m at most, beam of 0.1m for each individual hull (based on the 1200x600x100 XPS foam piece I've settled on using and general car related considerations), I've also already got the motors, and a fair few of the other parts. Most of the sensors will either be reasonably light (differential gnss, imus, cameras, lidar) or mounted underwater where they displace most of their own mass (em log, dvl, echosounders, modems) so I don't really expect significant extra weight from those. But yeah something like a manipulator arm, solar panels, landing pads, or more batteries would definitely load it down. Probably won't be any good for towing in any case with these thrusters, at least the last one struggled with that, haha. Definitely saltwater capable, but mostly calm-ish Adriatic waters and the occasional lake.

A few more parts arrived in the mail today and I've been able to weigh them and run some numbers again on how much the current setup should be in terms of dry mass:

• 1kg foam core for both hulls (~31 L of XPS)
• 2-4kg of epoxy and glass (very rough estimate)
• 1kg electronics box
• 2kg battery (4S4P or maybe 4S5P, 16-20x66g of 21700 cells, so about 1.0-1.6kg plus wires, nickel strips, main power switch, etc.)
• 1kg aluminium extrusions (4x125g crossbars, 1m side rails at about 2x200g)
• 1kg misc plastic and carbon fiber (masts, various 3D printed adapters and mounts)
• 1kg other cables and electronics, motors, etc.

Something like 8-12kg in total, but it is a very rough estimate still. To draw the waterline there it's about as such (ignore the current shape, was messing with the rear and it doesn't change the volume much):

An extra 10kg of payload on top of all the usual equipment seems like a reasonable max, but it would sit quite low.

seasquirt said:

fixed form or de-mountable, engineered grid with tapped holes for payload attachment

Click to expand...

Well the main approach I'm planning would be to have a very well bonded continuous aluminium rail on each side, with holes spaced 15mm apart for screwing through tapped 1515 V-slot extrusions. Then four or more crossbars can be mounted in any of the opposing holes, maybe with some additional bracing if need be. And anything compatible with the 1515 standard can then mount just about anywhere. How well will this resist twisting force from waves trying to tear the whole thing apart? Good question, I still need to decide on the thickness and order the side rails

MoffKalast said:

Well, when it comes to research use there is unfortunately no such thing. Projects come and go, one day you need slow bathymetric mapping, the other day you need long range high speed, the third day you need sideways docking, the fourth day it needs to just stay still in the middle of a lake with a large board so drones can land on it.

Click to expand...

That would be a constraint on the SOR. If it needs to be versatile, still there will be limitations. A 10 foot boat will not be possible to be deployed in a small pond. Conversely, a smaller 4 foot boat would not be adequate for the ocean. No design will do everything, so you have create an SOR that will define the limitations.

gonzo said:

No design will do everything, so you have create an SOR that will define the limitations.

Click to expand...

Sure, but I think I've outlined the general idea reasonably well so far, no? Yeah if I'm being entirely frank with myself, an inflatable paddle board or whatever would likely be a better fit for my strict requirements. Unfortunately "must look very cool" is one of them that I've forgot to mention so far.

MoffKalast said:

semi-displacement

Click to expand...

seasquirt said:

Look at fish

Click to expand...

So anyhow I've been working towards testing a semi-displacement option, and funny enough what I've got currently looks exactly like a fish from the side with the thruster extension looking like a rear fin:


Note to self, attach the two hulls together first before sanding next time so they're actually equal The XPS sands really nicely though, so making compound curves may not be such an issue. Tested out using wood filler on it and that is a problem since it sands slower and it's impossible to get it flush without ripping out all the foam around it. Gotta find some kind of foam filler I guess.

All assembled together:


I've tested out how it handles in different payloads and center of mass locations, since it's half scale, doing eight times mass scaling should be roughly correct given square cube law.

This is around the minimum mass (the waterline is fairly visible on the inner side):



This is really slippery in both configs, might have some proposing in high waves if loaded too far back but that's adjustable I guess.

And max mass:


Yeah I think having it weigh more than like 15kg total would make it really chug. With the rear CoM configuration the stern goes completely underwater when reversing and the mid one has a lot of resistance.

I've been running some more numbers on the composite part and I think if I switch to carbon fiber it should actually be a lot less than I initially ballparked, about 1.5 square meters total for both hulls at 200g per equare meter + equal mass of epoxy should be about 600 grams, not 4 kilos lmao. Depending on how much the electornics will be I think the 7-8kg figure might be achievable with no payload.

P.S. While researching semi-displacement hulls a bit, I found this test which really outlines what I was seeing on my previous build, i.e. the bow lifting out of the water entirely and the stern sinking in. That's mainly what I'd like to avoid with this one so I think this is likely to be the right direction.

Oh right, and that's how the cad model I've printed the cutting templates from looks currently. Deadrise of about 10 degrees, stern raises up about 45mm from the keel, slightly sharper front, about 30L total.

Alright, so I've had the chance to run some pull tests at a pool this week (pool pull tests?) loaded up at scale 10kg, and while I had way too much stern trim and didn't really manage to pull it straight I think it performed really well:


The transom design is probably not the best since there's so much turbulence there, but it was almost fully submerged. Still, I've set up the test model to have one really sharp flat transom and one where the curve continues after the motor mount, and the curved one does seem to produce somewhat less turbulence, at least at first glance (top one is curved):

(imgur.com https://i.imgur.com/ZrHE76D.gif in case it doesn't render)


Since the main problem seems to be loading this up too much for the beam, so it becomes inefficient in high draft conditions I've shortened the freeboard slightly and will be gluing extra 2 cm sections to the side of each hull, which adds a lot of buoyancy where it's actually needed. So overall I think this'll work decently enough.



Thanks for the input guys, I'll let you know how the final version performs

MoffKalast said:

.... and while I had way too much stern trim ...

Click to expand...

That is because you are too heavy for the length. Your length-displacement ratio is low, and as you approach the prismatic hump, she will squat. You can't avoid it.
Unless you increase the LD ratio...i.e, same weight, but double the length, or the reverse, much less weight on same length.

It would help if you Designed it first (ie drew a GA and did a full weight estimate), rather than a lot of trial and error.
The errors you are finding, would be easy to spot if you Designed it first....since the weight, is your killer and influencing its performance.

Bathycat 1 | Boat Design Net https://www.boatdesign.net/gallery/bathycat-1.30211/


Bathycat 2 | Boat Design Net https://www.boatdesign.net/gallery/bathycat-2.30212/


This can be an option for you.

Alright it took so long to get to this point but I've finally pushed it over the line and tested the first iteration.

I've thrown together an imgur gallery of some more photos of the build for those interested in the process, though I missed a few phases. It was just so much work haha. If I've measured it right it's about 8 kg in total with current gear, and gets about 7.5 kts at 90% throttle. That's as much as I've let it rip it until I mount more temperature sensors and a current shunt to make sure I'm not overdoing it.

Its first job was to film some AUVs underwater with a GoPro on a stick, and it could run circles around them, it might get to test out a sidescan sonar next, we'll see how that goes...

Ad Hoc said:

That is because you are too heavy for the length. Your length-displacement ratio is low, and as you approach the prismatic hump, she will squat. You can't avoid it.
Unless you increase the LD ratio...i.e, same weight, but double the length, or the reverse, much less weight on same length.

It would help if you Designed it first (ie drew a GA and did a full weight estimate), rather than a lot of trial and error.
The errors you are finding, would be easy to spot if you Designed it first....since the weight, is your killer and influencing its performance.

Click to expand...

Yeah the squat is actually really visible in the video too, when it really starts to churn foam, but I think it's not too much of a practical issue. With better hull trim, cruising at 5-6 kts it's already negligible and I still have to trim the motors so they push up a bit, then it should be even better (thrust is completely level with the waterline right now since it's just one on a long list of things to adjust). Ultimately this is the largest size I can do for transport and I would always add more weight if I could. A rigid-inflatable would've been a more practical option, like the WAM-V or something, but that's its own can of worms in terms of manufacture.

Idk, I like the trail and error, it's more fun. Seeing is believing as they say.

gonzo said:

Bathycat 1 | Boat Design Net Bathycat 1 | Boat Design Net https://www.boatdesign.net/gallery/bathycat-1.30211/

Click to expand...

Oh that one's a chonker, I love it

3 x T200 seems like slight overkill for the size, does the middle one pivot around for steering? What kind of payload and speed does it manage? I think just the thrusters on that cost more than my entire build minus the sensors haha.