Sail-Trim Simulator | Hacker News Reader

Sail-Trim Simulator

simulator.atterwind.info

140 pointsstass28 comments

Here is a fun sailboat race simulator. It's very simple, but pretty realistic especially for open ocean racing. You can use their default race course, sometimes join a planned regatta, or make your own. (It was great during COVID, to try racing across the oceans, or around the world. It's in real time, so you might not need to change anything for days If you are in the trade winds, and open ocean, but then when you get closer to land, things get "Western" very quickly - just like with a real boat. It's called SailNavSim. (Learned about it from HN.)

8bitbyte.ca

I am a lake sailer. And without landmarks or anything it is impossible for me to perceive the boat as moving.

Even some kind of wake or something would help.

I thought I was doing something wrong. It doesn't seem like anything I do has any effect, but guess that's just because there is no apparent effect.

Eventually it made sense that boat-speed only changes the "apparent wind", as it's only simulating wind-boat interactions, not water-boat interactions.

But to get the sail-wind interaction you need the movement of the boat. To get the movement you then need the hull-water interaction. Modeling just a static, vertical, sail doesnt really demonstrate anything practical. It is more akin to similating a captive foil in a wind tunnel.

The obvious next step would be to calculate the cog and tilt the entire rig in response to sail forces, which sheds wind until a balance point is reached, or the boat flips.

One of the most mind bending facts I tell people, even sailors, is that sailboats are not limited to sailing at the true windspeed. Sails are wings, not bags. In fact, a boat's top speed is directly dictated by its ability to point into the wind (assuming, for example, the boat is not physically limited by it's displacement hull speed, as in the case of hydrofoils). The consequences of this simple truth are manifest.

First, consider the edge case where the sail is acting as a bag when you're sailing downwind. As the boatspeed approaches the true windspeed, the apparent windspeed falls to 0 and the sail will luff. In this specific case, the boat can not go faster than the wind.

Now consider the boat cutting across the wind at a 90 angle. When the boat starts moving, the wind comes 90 degrees off the bow. As the boat increases speed, the apparent wind shifts closer to the bow. Apparent wind is just vector addition of true wind and boat wind. If the boat achieves the same speed as the true wind, then the apparent wind is sqrt(2) ~ 1.4x faster than the true wind. More wind means more power, so with that additional wind, it can go faster. Continuing the example, as the apparent wind increases, it appears closer and closer to the bow. Eventually the sail will stall and produce less lift. This is the point where the boat will go no faster.

The slowest point of sail is directly downwind. In a race, it is often much faster to gybe back and forth rather than ever go directly downwind. When a boat goes directly downwind, their boat speed cancels out the true wind. In the strangest case, if a high performance boat going faster than the speed of wind (say, on a broadreach) goes directly downwind, the apparent wind will appear to be coming head on. They've effectively gone 'into irons', yet they're facing 180 degrees off true wind.

If you ever get the chance, you should see the SailGP boats race. Their sails are almost always hauled fully in, even downwind. The other thing is that they gybe downwind because to go directly downwind would be to stall. In effect, these boats can achieve multiple times the true wind speed, but so long as they aren't pointed directly into, nor directly away from the wind.

Into the irons/in irons being a dead sailing area, where boat is head on into the wind (wind's eye). A lot of people get surprised that you can sail upwind as well.

A great project, but it needs some kind of grid movement and rocking, foam or something so you can perceive its movement.

If other people are interested in sailboat physics -- this resource is a goldmine of information on how sailboat and sails work and physics around it: https://www.onemetre.net/Design/Design.htm

You can set the wind to 0 and the speed somehow stays at 35 knots, I don't know much about sailing but that can't be right

Very cool demo...

Most yachts have polar charts, and the speed is a function of sail area, heading, wind speed and direction amongst other things. Are you considering calculating estimated boat speed based on given conditions and controllable variables?

Not my demo! But I got inspired by it some time ago to build a simulator for monohulls which uses a physics based VPP to calculate the boat speed based on input controls like mainsheet, jib sheet, travelers, backstay and so on at https://www.sailrhythm.com/

It uses Catalina 36 Tall Rig as a base model for sailboat parameters and was calibrated to match ORC-published polars for it within 1-5% on both close, beam and broad reaches.

Mostly built it for myself to help me understand how less common control affect the sail shapes, angles of attack and boat heel and behavior in a visual way.

It's still WIP, but you might find it useful for yacht-specific stuff!

What are you using for your physics engine, or how did you go about designing your aero model? I've been poking about with wing aero for a toy flight simulator and there's not a lot written about the design process of modeling this stuff in near-real time.

I cheated a little bit and don't calculate the fluid dynamics in real time. Instead, I'm using force models from the excellent Aero-Hydrodynamics of Sailing book by Marchaj, which is sadly out of print. I then do iterative balancing of the drag force and combined drag/lift forces on sails until they converge.

Little confused. So you're mostly giving credit to a demo that then inspired your own demo?

Either way, both demos are fairly interesting, especially since there's so little exploration of sailing and the physics involved most of the time.

The second sailrhythm demo seems rather high quality and a better representation of the physics involved. Not sure if you're planning to work on it further, yet putting the rudder heading up near the top would be helpful, maybe with a circular dial you can select like the compass or broad reach displays. Kinda weird when the boat flips completely around to head left.

Some kind of visual representation on the Boom Vang, Cunningham, Outhaul, Backstay, and Jib Lead would also be helpful, since it's really difficult to tell if they're actually doing anything. Maybe a transparent deformation magnified overlay showing the sail profile surface deformation extended away from the actual sail in extreme distortion. Otherwise the changes are so minute its challenging to perceive.

Otherwise, it's a cool demo, and seems to be (from a not especially experienced sailor) relatively realistic of the forces involved. Suggested extension would be showing the wind flowing off the sails and how the forces transition. Also, showing other boats, particularly something complicated like a Windjammer [1], Full-Rigged Ship [2] or Schooner [3] varieties.

Also cool, simply because so few submissions seem to deal with sailing, boats, and generally water transport principles. Lots of possible VC ideas related to sails and sailboats. MarineTraffic [4] is currently tracking 300,000 ships out on the world oceans, and surprisingly small number use any form of sail based propulsion. Route planning and "this much cargo to this area with this speed of delivery" estimation relative to money spent on oil and fuel for a normal ship would likely be possible. From this Quora answer (suggested by Google) a small container ship (something with 15000 kW, 2000-3000 TEU capacity, maybe a Feedermax ship) might use 2000 tons of fuel crossing the Pacific (maybe $500,000 to $1,000,000 of fuel) [5]. This paper [6] has a much more complicated breakdown if you're interested in those types of calculations.

Also quite a bit with drones, and automated sailing. [7][8] Even a decent amount with land propulsion [9] that's been explored occasionally and there's even a tiny amount with space probes and planetary rovers as concepts [10].

[1] https://en.wikipedia.org/wiki/Windjammer

[2] https://en.wikipedia.org/wiki/Full-rigged_ship

[3] https://en.wikipedia.org/wiki/Schooner

[4] https://www.marinetraffic.com

[5] https://www.quora.com/How-much-does-it-cost-to-fuel-a-cargo-...

[6] https://www.man-es.com/docs/default-source/document-sync/pro...

[7] https://en.wikipedia.org/wiki/Unmanned_surface_vehicle

[8] https://www.saildrone.com/

[9] https://en.wikipedia.org/wiki/Land_sailing

[10] https://en.wikipedia.org/wiki/Zephyr_(rover)

SailRhythm is a bit WIP, so didn't want to share too early!

Thank you so much for the detailed feedback. I do have plans to implement wind flow visualization and have a way to visualize forces as well. Controls a bit tricky to do, but this is definitely something I've been pondering. As you said, at certain headings the changes sometimes are too small to perceive.

You're welcome on the feedback, and mostly just suggestions from trying the product / demo. Cool generally that somebody's actually investigating sailing.

Another thought of (if there's actually further interest in this kind of software) would be middleware like SpeedTree (https://en.wikipedia.org/wiki/SpeedTree). Seems like sailing games have been pretty popular lately (Sea of Thieves (40m player base), Assassin Creed IV (15m)) so there's probably at least some market for modestly priced middleware that handles sailing physics.

The blue cones are probably the wind coming to the boat but there is a wall of cones a little further away, visible by zooming out. Those cones don't have the same directions of the ones close to the boat. It's probably apparent wind (close to the boat) and real wind (further away.) It would help to start with a zoom level with both winds visible.

The biggest problem that is puzzling everybody is that speed never changes. I eventually decided that this is because the app calculates the setup of the mainsail to get the desired boat heading and speed given the wind. If you keep the heading constant and change the boat speed you see that the sail rotates around the mast.

It's the opposite of what we do when sailing: we set the sail to a shape and direction (let me use these terms) to go somewhere. In this simulator we do it backwards and adjust the wind to get the sail into a shape and direction.

Thanks for creating this!

I am still learning and would love to use this, so I can validate my thoughts :D What would be great: - change model. I sail mono hull and would love to have - like others stated. Som kind of indicator that we are moving - or clouds

That wake thing would've totally helped me too, felt weird not seeing any movement. Glad I wasn't the only one who got thrown off.

I like the idea but I don't get this at all. Shouldn't the speed change based on the mainsail trim relative to the heading? I pointed the boat dead downwind, eased the mainsail all the way out and the speed is 0 kn. I can't even see the value for the trim, just speed, but you get speed from correctly trimmed sails.

why so heavy on the water graphics and light on the sail?

You get water graphics for free from three.js

Pretty awesome if you ask me.

Why even use Unity if this exists? Why even install an app if you could feasibly run a full-blown game with 3D physics in the browser on your iPhone at 60 FPS? Where in the world are all the browser-based games?

It would be great if it would be able to limit the max speed based on the heading, right now the slider allows the same max speed for any heading. On this boats I suspect max speed is achieved going on a beam reach (90 degrees to the wind)

it's hard to tell if you're moving

I guess I’m not the only person who opened the link from yesterday and didn’t see anything about sailing.

https://news.ycombinator.com/item?id=43756926

Thank you, you’ve resolved my despair.

Reply on news.ycombinator.com