Close

Let's just do it and see where it ends..

A project log for Reflecting the sun into your home

My home doesn't face the sun, so it's dark and cool inside. I want to use mirrors to reflect sunlight into my home.

mimemime 04/10/2015 at 14:290 Comments

I've thought a bit about the three strut construction, and how and why it would work. What it came down to was this: there are three axes of linear translation and 3 axes of rotation. If not everyone of these axes is constrained, structures will not be stable. That's why for existing heliostat structures we see something that, for instance, is fixed in the translation directions, and can vary in rotational directions. For instance, think of a big satellite dish, a common construction is one that can rotate around its base, and also in the vertical plane (also how many heliostats work).

However, I don't want to make such a structure. I'm intrigued by this structure. I want to see how well this would work. The trick is that instead of using universal couplings for all strut points, this structure is fully constrained by constraining the struts at the bottom. No X,Y, Z, or rotational movements allowed. (Well, except for one of course, in the horizontal plane.)

The mirror corners are constrained in X,Y,Z but in none of the rotations. The mirror will then be (loosely) constrained by the elasticity of the strut threads, M6 metric thread, which is fairly elastic. Basically, this structure is the inverted situation of what you would get if you hung a mirror from the ceiling with three ropes. The only difference is that gravity is not really acting as a stabiliser in this structure.

Will it be easily scalable to larger structures?

Not really, I'd say.

Will it be wind proof?

Well, let's say that it will behave more like a bundle of reeds.

Will it look elegant?

Well, I think so. All the active bits will be housed in the base, and the only thing that will be seen is the mirror, three struts, and a base.

Will the settings be repeatable?

Well, this was my fear before, that it wouldn't, but I decided that unless the M6 threads will be bent too much (beyond their rated stress), they will not permanently deform, and they will operate nicely in their elastic region. In that case it's just a play of the same repeating forces, so they should result in the same positions. I'll just have to make sure they're not bent too much. How much bending is allowed remains to be seen.

Construction

In the base there will be a ball bearinged sleeve*, 19 mm high, and underneath a motor with a gear reduction. The sleeve is there to prevent off-axis stresses on the motor. Motors don't like too much stress from unintented directions, just as humans.

When the motor turns, the M6 thread turns too, and the nut, which will be prevented from turning in some way, will move up and down. The ball end joint is only there for structural support, really.

I've taken some more measurements, and in the end it transpired that in order to catch sunlight in the morning the mirror is almost vertical. Hence the ball end joints. It all depends somewhat on the angle of the ball end joints on how extreme the angles can be. I have to be careful not to 'over tip' the mirror anyway (up side down), because that can destroy parts of the mirror once it starts moving in to the wrong angle.

What Motor Size and Rotation Speed?

M6 has a pitch of 1 mm, so for every rotation, the thread moves one mm. If the mirror has to be set every half hour initially, then the strut has to move up/down with about 7 cm. That's 70 mm. So in order to not wait for too long (less than 10 seconds, I'm an impatient man), the motor has to turn with at least 7 rotations/second, or 420 RPM.

The current mirror weighs about 2.5 kg, which is roughly equal to 25 N, assuming all the weight will be on one strut (which is not the case, but we're doing a worse-case calculation). The shaft is 6 mm in diameter.

I've used this nifty online torque calculator to calculate the torque required:

it's 0.0115 N.m for raising and -0.00354 for lowering (using standard friction coefficients).

So unless I've got a gear reduction motor or a brake, or if the motor is feeble in general, the struts are going to slowly lower themselves, which is not what we want.

... I'm off to find some geared motors on ebay.

* new words added to the English languague: bearinged, i.e. the state of being equipped with bearings.

Discussions