When North Sails released their video on design, starring Michael Richelsen above, I thought it was a bit over the top. But then again, it’s hard to explain how much time, effort and money that goes in to all those tools.
Being both an avid sailor and an engineer (M.Sc. Computer Science) I want to dig deeper to really understand what’s going on. When North introduced 3Di, I went to Nevada to try to explain why this could be the biggest thing in sailmaking since 3DL. Al major websites republished the article and it even ended up on Norths owns sites as “3Di Explained”.
With the big projects like America’s Cup, Volvo Ocean Race and MedCup, sailmaking have gone from being an art to much more of a science. And if there’s one person who really understands this development it’s Michael Richelsen, the man behind much of the North Sails Design Suite (NDS).
Michael studied mechanical engineering at the Technical University of Denmark, and wrote his master thesis on sails and aerodynamics, before joining North Sails in 1979. He’s one of the leading experts on CFD (Computational Fluid Dynamics) and he’s been involved in America’s Cup projects since 1983 and he managed the performance prediction group for Alinghi in the last Cup:
What makes this boat so different to, say, a keelboat is that windage is much, much bigger. Windage goes up by speed squared, so when you think that this boat is capable of travelling three or four times faster than a Version 5 boat, you can see the importance of windage. With the keelboats, wave drag was the biggest element of drag, where at a certain speed the boat ran into this wall of resistance. There is no such wall with this kind of boat; that aspect is not there. Aerodynamic drag is a much greater factor, and you could say that what’s going on above the water is much more important than what’s going on below the water this time.
North Design Suite. There’s many pieces to a complex puzzle. We all know how everything interact in the real world, and it doesn’t become any easier trying to do it in software.
What’s happening right now?
Following Alinghi loosing the AC33 I returned to my full time job at North Sails. Having been on reduced time for quite a while, the North designers had accumulated a fairly long wish list of desired upgrades to the North design suite of software. The development of the software falls in my department – fortunately we now have a couple of great guys helping out on this. High on the list was accurate dynamic modeling, i.e. upgrade the current static coupling between Flow and MemBrain to account for external variations in time due to wind and/or sea state. Collectively we decided to attack this first, but part way through the development the wing for the next AC34 entered the scene.
We decided to shelf the dynamic work for the time being and instead focus on adding a wing model to Flow and MemBrain. Early on in this phase, North Technology Group, was contracted by ACRM to design the AC34 wing offered in their Shared Design Package.
So the wing development combined with joining our small ACRM/SDP team has kept me busy!
What’s you sailing background?
I started sailing with my dad and then joined the junior program in a local sailing club. Grew up in the classic “juniorbåd” followed by Yngling, then started sailing Trapeze dinghy and 470. Also sailed in various Danish and Scandinavian one design keel boats, after I finished school and began my studies.
How did you end up at North Sails and in CFD?
I studied Solid Mechanics at the Technical University in Denmark with an emphasis on applied mathematics, including Fluid Mechanics. For my M.Sc. I chose to create an aero-elastic model for a sail. As I did not have access to aerodynamic data, I wrote to North Sails San Diego (back then the Mecca of North) to inquire whether they would be interested in collaborating for my thesis work. Lowell North invited me to come to San Diego and then go down to NZ to work with Tom Schnackenburg for half a year, who became a kind of mentor for me. That was the beginning of it.
I went back to Denmark to finish my M.Sc. Following that in order to continue my work with North, the company offered to sponsor my Ph.D. on further aero-elastic development. Our structural MemBrain model originates in this work.
Flow applies wind pressure to 3-dimensional sail mold and provides sail forces. Here’s a predicted pressure distribution and sail wake.
How were you involved in the design suite?
When I first came to North, daily design work was based on 2D patterns. Schnack was pioneering a 3D ”TinSail” program, which ran on a main frame computer (those were the days) and not easy to use. My task then was to spear head the development of a PC based 3D CAD suite for designing and building sails throughout the whole company. This involved 3D shaping and derived panel layouts and broadseams. Later on 3DL and 3Di was added to the company inventory and brought into the scope of the design suite – still based on the same software concept. As this CAD suite matured, and PC’s became more powerful, we moved on to the analysis package consisting of Flow and MemBrain plus the VPP.
How close to reality do you get with those tools?
In my recent AC campaigns with Alinghi, all aero load cases used for the structural design work on the boats have been based on Flow / MemBrain predictions. Over time we have come to trust these predictions to such a degree that these tools are used extensively for custom projects. Secondly, all North 3DL and 3Di layouts are based on analysis using these tools.
MemBrain balances wind pressure, sail shape and rig forces. Comparison of sail in MemBrain and on-the-water photo for a TP52 jib.
As I’m going to University of Auckland next week to visit the Twisted Flow Wind Tunnel, could you say anything on the relationship between practical experiments and CFD? In theory? In practice?
I have been involved in wind tunnel tests at various occasions. I find the tunnel a very efficient setup for investigating a broad range of sails – and at the time highly visual. Therefore, it will be hard not to learn something from a tunnel session. The difficulties in a tunnel lies in the problematic scaling to full size plus setting the proper twisted wind profile. You need to be extremely meticulous in your testing procedure to achieve consistent results – otherwise you will pull your hair out later trying to make sense of the data. In my opinion neither tunnel nor CFD can replace one and another – both are great tools in their own sense, each with their strength and weaknesses.
Can you describe the North VPP?
The North VPP is modular in its program structure. This means that the predicted performance is based on a given set of building blocks, each representing a specific component of the combined boat, rig and sail model. For instance one block can contain the hydro force and moment components measured in a towing tank. Alternatively the same hydro block can consist of multiple CFD blocks, each representing a specific hydro component, such as canoe body, appendage(s), etc… The greater the number of blocks the greater the level of detail in the resulting output. Same goes for the aerodynamic components – windage for instance can be accounted for by a single, overall block or broken into individual components.
Blue dots in charts represent measured on-the-water performance data from TP52 Platoon in all races she won in 2008 MedCup. The black dots are North Sails VPP predicted performance. Also, check out Chris Wilsons article “North Sails VPP Action”>.
Furthermore, the North VPP can find a solution for various degrees of freedom from 1 to 6. For normal displacement keel boats we mostly run with 4 DOF with assumed equilibrium in Heave and Pitch – corresponding to testing in the tank with free sink and trim. For lighter displacement boats, in particular multihulls, we use the full 6 DOF, as these boats are more sensitive in sink and trim to variation and position of ballast, such as crew.
For the ACRM wing, however, we also ran the VPP in 1 DOF – only enforced Roll moment balance – in order to optimize wing aerodynamics without the plat form underneath.
In other words you can tailor the VPP to match your needs depending on the project at hand.
Das Boot. A hull being run through a series of heel angles to produce a matrix of hull drag to be used in the VPP.
What’s the biggest challenge in the next few years?
We have begun to move beyond steady state to dynamic analysis.
Secondly, we have started to use RANS codes to a larger extent as a more accurate CFD tool than Flow – in particular for downwind sail design – in what we call the Virtual Wind Tunnel.
Now you’re working on the wings for AC34. How different is a wing compared to a sail from a CFD perspective?
Same Flow / MemBrain model as above now incorporates an aero-elastic model for a multi-element wing plus headsails. This way to us the wing is simply a rigid (and thick) type of main sail with additional trim controls. The ACRM wing design by North Technology Group is designed with this package with the addition of Abacus for structural detail work and OpenFoam for high lift CFD.
What’s the biggest challenges with a wing?
Keep the cat from flipping over, which won’t look pretty on an AC72.
Wings have already been developed over quite some years on for instance C-class cats. AC wing designs start from here, and they will go beyond the current norm, though much is dictated by the AC race course format. I expect we will see smarter control systems than on C-class wings – also to handle the larger power of the AC72 wings – and sleaker aerodynamics. Windage is huge on boats this fast.
Virtual Wind Tunnel models air flow on upwind and downwind sails. This is a run of a Ker 39 medium jib and main showing detailed analysis of the hull and sailplan interaction
Do you get to race anything yourself?
Not these days, as last summer I moved myself and family to Bend in Oregon, USA. Instead we have just started on a full season of skiing.
Any career advise if you’re at the university and are excited about the science behind sails and wings?
If you study at a university and are excited about the science behing sails and wings, engineering would be the way to go. Here in my experience there are two distinct paths to follow, depending on whether you aim to be a designer or developer. As a designer you will be using the tools with your scientific background to design the best possible product. As a developer you play more of a supporting role in developing and maintaining the designer’s tool set. You are not designing the physical product per se, for instance a sail, but rather a more abstract product. Both positions can be equally satisfying – so base your choice on your interest.
Due to the work load I cannot imagine someone covering both positions, and consequently I do not really design myself. But I do enjoy a close collaboration with the various designers within the company, which I find highly stimulating. In terms of what course work to choose at a university, based on my personal experience the developer path will require an emphasis on applied mathematics in areas of numerical analysis and modeling, both structural and aero/hydro-dynamics. From my perspective our designers are more broadly based engineers – or with a naval architect background. By mean of their jobs they do get to go sailing more often than I do ….
All to say that you should pursue your personal interest. And make sure you truly enjoy sailing (and racing), because in our business you inevitably end up tinkering with – and talking about – boats, rigs and sails for the larger part of your life.