Automating the Bedroom Window
Master Doc.
(Draft in progress; please contribute)
Goal:
Fully or partially automate the raise/lower of the bedroom read window that's positioned behind the bed. If only partially lowered, consider adding a thicker strip to the top of the window to serve as a sort of railing.
Challenges:
- Window is approximately 80 kg
- Window has an arclength (overall length) of 6 meters which potentially leads to large overhang moments
- Blending 3D contour of Seapod shell with 2D contour of window.
- Sealing from the elements and not compromising R value
- Hiding components allowing access
- Automating affordably, ideally with off-the-shelf products
Concepts (newest to oldest):
- Finalized Hybrid Concept
- WORKING CONCEPT: hybrid concept - see below
- Using only linear actuators on the ends - see below
- Gilles' working concept as of 2-13-21 (follow link)
- original video concept
Finalized Hybrid Concept
Working Hybrid Concept - notes and scratch
Gilles' comments about our new paradigm recorded here:
I agree with him on may fronts. This is exactly how we should be thinking about the center of gravity. Here is a diagram to help quantify his point about the center of gravity. There are two forseeble locations we could place the lifting locations.
Option 1: places these within the headboard of the bed, about 2 meters apart from each other and 1.5 meters from the end of the window. This option will be easier to drive off of just 1 motor and connect the wires between the two lifting mechanisms.
Option 2: places these in directly in line with the center of gravity, 3 meters apart from each other and 1 meter from the end of the window. This option will have benefits from a balance and moment loads perspective. It will be difficult to connect the two lifting mechanisms in this case, and they may need to be separate systems entirely.
Option 3: we could potentially build the bedside table in a way that allows us to split the difference between option 1 and 2, and still hide our sins behind it. A table with drawers could be designed to offer access to the mechanisms.
As for adding lifting mechanisms in the wall, to lift the window near the center of gravity, Gilles also has a good point. The problem with using linear actuators for this purpose is that linear actuators usually have extra length to them beyond their given stroke. For instance, a linear actuator that moves 600 mm, might be 750 mm long total. This is a problem for us because the wall-height is only 670 mm, and we need at least 600 mm of stroke. Therefore, most linear actuators I've found with enough stroke would stick out of the top of the wall. E.G:
Alessio said this problem can be alleviated if we bury the actuator into the floor by a few inches. Waiting for confirmation on how many inches is acceptable. Here is an example of such an actuator.
The counterweights should be retained. We can design the counterweight system, much like our original concept. Only, instead of using a brass pole, we use square box-tubing and secure a cheap linear guide to the side such as this.
Then we would have the benefits of precision engineered bearing system, with the simplicity of the original pole-counterweight system. All of this beingrelatively cheap to source. Also, compared to the system below that uses "only linear actuators", these linear guides can be lighter-duty and cheap. This is because we will be supporting the window closer to the center-of-gravity, and the moment loads will not be near as extreme as in that last concept.
Alternatively, we need something lower profile to drive the the window near the center of gravity. Perhaps a pulley/cable system in the headboard of the bed. I noticed the mechanics that control the slide-out function of an RV slide-out. Its a combination of linear guides, pulleys, cables, motor, chain-drive, and mechanical advantage, albeit in a 'horizontal application' . These mechanics are cheaply available only. I think we can hide something like this in the headboard of the bed/wall to drive this window up and down.
Altogethor now, something like this:
Attempting to raise/lower the blinds as well:
Idea would be to include a similar mechanism on the front face of the aluminum extrusion to lift a rigid shade.
Issue is that if the shade has the same stroke of 600 mm, the rounded edge would not lower entirely into the wall and that portion of the window would be permanently obsucurred. Exploring other ideas.
Using only linear actuators on the ends - After reviewing with the team, this concept is too expensive and perhaps over engineered. We have pivoted towards adding support underneath the window, within the wall. See hybrid solution above.
This is an attempt to use a linear actuator (self-contained unit) to solve the problem. While these can be expensive, the tradeoff will potentially be in time spent engineering, proving, and troubleshooting a custom solution. Therefore, this can also be thought of as an "easy-button" to solve the problem, which is not inherently an easy problem.
Considerations:
Drive type:
-
- Ball-screws vs. lead screws
- Ball screws ~90% efficient - > can backdrive
- lead screws ~ 40% efficient - > usually can't backdrive do to inefficiencies
- Ball screws are more precise and expensive
- lead screws sometimes use a plastic carriage rather than a steel
- Belt drive seems more oriented towards horizontal applications, whereas ours is vertical
- Ball screw is most likely the most appropriate for our application
- Ball-screws vs. lead screws
Lubrication:
-
- Screw drives and linear bearings usually require periodic lubrication
- Lubricate using grease rather than oil for vertical applications
- Some ball screws will advertise that they are maintenance and lubrication free (internally lubricated)
- These products have a finite lifespan. Once the internal lubrication runs out, the product may fail.
- Depending on our intended # of cycles and lifetime, such a product may still be appropriate.
- Some lead screws advertise lubrication free
- plastic on metal
- Likely not suitable for our heavy-lift application
Linear guides:
-
- will help align the window pane and resist the overhanging moment loads (which are really heavy)
- therefore, these must be robust due to the high pitch and yaw moments (392 Nm)
- For instance, a counterweight or screw-drive system on its own would be unable to counterbalance these moments, so we need robust guide rails to do so.
Travel life:
-
- Travel life seems to be an important factor:
- Assumptions for travel life. Raised/lowered once per day, everyday for 100 years
- 605 mm of travel (1300 mm per open/close)
- 48 KM traveled in 100 years worst case - on the very low side for these sorts of applications
Electronic Brake:
-
- Some systems allow electronic emergency brakes. These are electric open (24 V) - spring close
- In other words, these are energized during actuation of the window, and normally stay closed
- This can alleviate the need for a counterweight and/or other safety systems
Linear actuator and motor calculations:
https://docs.google.com/spreadsheets/d/17lEcmTDpThAO0ecSoPah9yGxuAAONtuhxkV19Jpn0XI/edit?usp=sharing
Selecting a linear actuator:
https://www.lintechmotion.com/products2.cfm?ModelNo=170&t=Group6
Selecting a motor and gearbox (see calculation page):
https://www.amazon.com/Gyheung-Extruder-Stepper-Planetary-Gearbox/dp/B08LH19VBG/ref=sr_1_6?dchild=1&keywords=gearbox+5%3A1&qid=1614476207&s=hi&sr=1-6
Costing (TENTATIVE - big ticket items only - hold off ordering anything):
| Component | Make | Model | Price | Qty | Ext. Price |
| Linear actuator | Lintech Motion | 1741224 | $3,319 | 2 | $6,638 |
| 24 V Motor | Newark | DB59S0124035-A | $108.52 | 2 | $217 |
| Gearbox | Gyheung | B08LH19VBG | $50 | 2 | $100 |
| Approx. Total System Cost | $7-$8k |
Other notable systems:
https://www.igus.com/info/page-19897
-
-
- Could offer interesting options for counterweighting
- This model is unable to resist heavy loads (plastic on aluminum)
- Still unsure if a toothed belt drive is appropriate for vertical applications
-
https://www.youtube.com/watch?v=lQWBI0gnhpI