In my previous post I went, at some length, through the design process, but that’s certainly not the end of the story. Before committing needle and thread to fabric, I want to be confident that what I end up with is going to work. This is where prototyping comes in.
Prototyping is no more than making a straightforward / inexpensive model of the build to validate the design. There are a number of advantages to this:
- It shows that a design drawn in 2D on paper or in a CAD application works for real in 3D
- It helps assess whether the build will give me what I want
- It gives me the opportunity to change things if I don’t like them
- It gives me the opportunity to plan how I’m going to put the build together – ie in which order I should assemble the components
- It gives me the opportunity to practice the skills needed on inexpensive materials
- As it’s a relatively cheap exercise I can repeat it as many times as I like until I’m happy
It will be pretty obvious that there are no significant downsides to building a prototype – just a bit of time and a small monetary outlay. But that time and money could prevent me wasting much more time and money if I’d gone straight to the main build and got something wrong.
On a personal note, forcing myself to prototype forces a more patient approach than simply plunging into the main build. I know that I’m likely to have second thoughts about certain aspects, so forcing a slower pace and a bit of delay can give the space for that to play out before the main event.
Prototype 1: Paper
The first prototype is simply the design transferred to paper at an appropriate smaller scale. This helps check the measurements and the geometry. I end up with something small that’s quite fiddly to put together, but it does give me a first glimpse into the overall shape of the shelter.
The paper prototype is also a good way of planning how I’m going to cut out the shapes from the actual roll of material. I create a long piece of paper at a suitable scale to represent the suppliers’ material widths (in this case, I drew widths for 1.1oz Silpoly, 1.1oz Silpoly XL, and DCF). I then simply cut out the design at the same scale and arrange them on the material width plan. The idea of course is to get them to fit efficiently, but also plan the orientation in terms of the weave of the cloth. Enough space needs to be allowed for seam allowance etc, and a bit of extra length of material for contingency.
You will notice that I’ve assigned letters to each panel – this is to help me join the right panels together.
When I’ve done this and recorded the cutting layout, I can then use the tent shapes to build the model itself. In this case I didn’t bother with seam allowance etc and instead merely sellotaped the joins together as closely as I could.
At this scale (1cm on model = 20cm of real thing), it’s fiddly to join, and using sellotape means the sides don’t necessarily hang at the right angles: the beak in particular. But it’s enough to confirm that geometrically it works (which it should do given I’ve used CAD to design it).
The keen-eyed reader will see that the original shapes drawn out had 3 different sizes for the vent cover. I wasn’t sure what size I’d need, so purposely drew several and tried each. I ended up going bigger than I’d originally drawn in the CAD version.
At this point I also noticed that the gap at the top of the doors that I’d left for the vent looked way too small, and trimmed a bit off the top of the doors. It still looks a bit odd because the vent cover isn’t quite right.
As mentioned above, if I stick with this design I definitely need to lower the top line of the doors to make the vent bigger. But I’ve also been having thoughts about the doors generally.
I’m pretty happy with the back and sides of the tent, and the main floorplan. When I look back over the iterations of design, it is pretty obvious that they haven’t changed much – all along it was really just a matter of finding the optimum balance of dimensions: I already knew the size and shape of inner I would want to use.
But the thing that consumed most time and resulted in several different designs was the front of the shelter. I played around with various door arrangements, some with a beak and some without. Although I started out going for a summer + tent, I very quickly got sucked into trying to make the shelter as bad weather resistant as I could. Ultimately, what I ended up with was some sort of amalgam of the two approaches. And having made a first prototype, I’m not that happy with what I’ve got.
So, before moving on, I decided to return to the CAD program and play around with a few options for the front of the shelter – essentially to see if there was any mileage in turning a complex hotch potch of ideas into something (or somethings) simpler.
I’m most concerned about the beak and the limited door height this causes: door height = size of space to crawl through on entry and exit. I learned this lesson with my short-lived experiment with a Trailstar, and to cut a long story short I’m not going to compromise on having a reasonable amount of height to enter and exit. So I decided to play about with various combinations of beak/no beak and flying/not flying doors to see if I might be happier with less of a compromise, albeit at the cost of not having as much of an allrounder as the end result.
My first thought was to simply keep the back and sides of the tent the same and just play with the doors. But ultimately, having a beak results in reduced entry space, and this is made worse by incorporating a vent too.
The answer came to me while I was lying in bed: raise the beak a bit. But I don’t want to overdo this. A further thought came during the day today, and I started experimenting with an 8-sided design rather than the previous 6-sided.
This actually looked ok after laying out the basic shape, starting with a regular octagon. I then cut the front half off and re-aligned the back and sides to follow better the shape of my inner (this hasn’t changed). Keeping the asymmetric idea from before, I then added a very short side panel in front of each main side panel, but turning in towards the centre slightly. This forms a better protected area under the beak. it felt like I was on to something.
Next I added the two inverted triangles that form the beak, using the same approach as originally – square it off at the front, and find the right angle to drop down at by trial and error. The beak top line is then a compromise of length, angle from the top of the shelter, and it also needs to work for attaching a front guy.
The extra panel at the side then gave me an idea – instead of attaching the doors at the front of the beak, I could do it at the back of the beak. This gives recessed doors, so when rolled up they’d be secured under the beak. I quite liked this.
Drawing the doors themselves as triangles rather than the previous trapeziums gave an interesting “V” at the top, and I realised that this could give me a vent. Because this is itself recessed, I could have a large under cover vent without adding anything externally to the shelter. Moreover, I don’t need to brace the top of the doors as I did before. They simply end up a bit like a wrap-around top garment – think dressing gown.
From the CAD layout, this also had the look of working if I wanted to “fly” the doors too, or even just have one flying. The possibilities seemed pretty good.
Excited about this, I built a second paper prototype.
Prototype 2: Paper
As before, I scaled everything down at 1:20.
Once again, I cut the shapes out and lay them on my fabric roll diagram to capture how much material I’ll need. This design looks like it needs an extra metre.
Incidentally, this time I wrote down the area of each panel, totted them up and came out with 7.4m2. Adding 10% to allow for seams and stitching gives 8.14m2. So I’m able to estimate the approximated weight of the fly:
- in 1.1oz Silpoly (1.24oz with coating) it’s 42g/m2 => weight of 342g
- in 1.6oz Silpoly (1.77oz with coating) it’s 60g/m2 => weight of 488g
- in 1.0oz DCF it’s 34g/m2 =>277g (and so a lighter DCF would be even less)
None of these weights, though, include fixtures and fittings, so they would be a good 100-200g more at least I reckon. But I’m more interested in functionality than the absolute lightest shelter, and even if it creeps above 1kg with the inner, I can live with that.
The next thing was the fiddly job of sticking the bits together. I had to check a couple of times when it came time to add the doors to make sure I was doing them right.
I’m much happier with this second prototype, and even though the shelter’s a little bigger, it’s still only about the same size as a Duomid, but with more sleeping space, and a better and more flexible front of shelter. This feels like a much better design.
The photo of the model doesn’t really show the amount of beak overhang protecting the vent, but it looks ok on the CAD version. It still may be that the vent isn’t protected enough, but I’ll have to see how it looks when I build a larger scale prototype.
Although I’ve gained a much easier vent system (hopefully), the downside is that the doors now attach in the seam between the beak and the extra side panel, giving me a three-way seam to deal with. I still feel like I’m ahead slightly though.
For completeness, here are some screenshots of the revised CAD design.
Happy with this, it’s now on to the next stage of making a full size prototype, which I’ll cover in the next post.