30.7.17

Building the Octagopn Pt.66 Dome shaped roofs: Perspective from below?


What about three panels rather than two? One above the other for an even better approximation to a hemisphere. Smaller panels are easier to handle working alone. If I was going to horizontally brace the larger panels then the 'bends' could easily become those braces.

Placing the two dividing bars/bends 30 degrees vertically apart ensures equal height of all panels. Will this overcome the perspective problem when [normally] seen from below?

Note how high a viewpoint is required for the upper panels to even become visible! Let alone providing any substance to the structure.

Only from upstairs windows would the top panels become visible. The visual effect from the ground would probably be one of a flat topped, faceted two ring structure.

A dome would be visually acceptable because the eye is easily tricked into believing the curve of a ball, or hemisphere, goes on around the back. Even if the top and back, were chopped off flat it would be completely invisible from below. Which is the normal viewpoint.

__________________________________

With the 2 panel high geometry, the height of the structure can be radically changed. As can the relative bending point between the two rings of eight panels. Providing the panels are still tangent to the hemisphere, around which they are bent, the upper or lower panels can gain visual prominence.

What it needs is equal perceived depth, of upper and lower panels, when seen from a chosen viewpoint angle below the roof structure. In my case upward viewing angles between 15° and 20° are appropriate. The equal visual height condition is easily checked with a drawing.

The image above shows a mock-up using PVC drain rods for the arc of the dome.  The canes in the foreground have been marked A & B. To get the upper roof to look balanced with the lower sloping "walls" A should equal B seen from the average viewpoint below the observatory itself. Which is looking 15-20° upwards in my own case. At the same level, or higher viewpoints, then another completely different ratio of triangle panel height to trapezium panel height will be required.

The equal panel height, which I modeled in card yesterday, has to be seen from about 45 degree or above to actually look visually equal. From below the ratio of panel heights can reach up to 5:2 or even 2:1 between upper and lower panels to actually look the same height form below.


Here is a drawing of a 'bent' cone with a 1:2 panel height ratio. The earlier 45° equal panel height model looked too 'weak' on top when seen from below.

COS 30° = 0.866. In a 10' dome = 8.66' [small circle] diameter. 
8.66' x Pi/8 =  Circumference of the small circle divided by the number of sides.

27.21' / 8 = 3.4' wide at the joints between octagon points. 
[Between upper triangles and lower trapeziums.]

Now I have to make another cardboard model to check how it looks from a typical garden viewpoint of 15-20° below. In fact it took two models to convince myself that 30° just won't work. Both roofs were far too flat. I was expecting a tear drop. And got quite a few. Of laughter! Next I'll try 40°. Wish me luck!



Click on any image for an enlargement.
*

Building the Octagon Pt.65 A dome by any other name?

*

The results of my earlier calculations in card model form were exactly as expected. I used a 3.9 height for both triangles and trapeziums. 3.9 for the trapezium base width and 2.77 for the triangle base width.

In plan view the choice of symmetrical dimensions produces a quite well-balanced, pleasing appearance.

Unfortunately this view is never seen except from satellite.
 As the eye level falls, the 'walls' quickly become far  more prominent relative to the triangles on the "roof" or top.

Only a drone or chimney sweep enjoys this view.

I was trying to think what it reminded me of.  [Spl.Inf!] A medieval royal tent?
When the eye falls close to or below the rim then the top triangles almost vanish from view. This is the normal view from the house and garden.

I think I may have to look at alternative ratios to reduce the prominence of the more upright 'walls.' The last image [right] is very similar viewpoint to the barn. [Left] I think I must have automatically made the upper triangles taller to achieve an aesthetically, more balanced appearance.

Since this is the normal view the appearance really ought to be considered. Upside is [perhaps] easier provision for an observation slit on two vertical panels.

What about three panels rather than two? One above the other for an even better approximation to a hemisphere. Smaller panels are easier to handle working alone. If I was going to horizontally brace the larger panels then the 'bends' could easily become those braces.

Placing the two dividing bars/bends 30 degrees vertically apart ensures equal height of all panels. Will this overcome the perspective problem when [normally] seen from below?



Click on any image for an enlargement.
*

AWR Stepper Drivebox Firmware IC Upgrade

*

Last evening's drives trial reminded me that AWR Technology had sent me a firmware upgrade chip a while back. 

The new 40 pin chip has the version number written on a label and a white spot to indicate orientation. The spot goes towards the long edge of the PCB. 

How difficult could it be? I collected a useful toolkit together and worked on a window ledge for better light. The weighted rubber hammer stayed in the workshop.

 Getting the original chip out was the easy part. I used a small, flat bladed screwdriver at first and then employed a larger one. Making sure that the chip rose evenly required gentle levering from each end. A 'boundary hedge' of yellow components near the edge of the PCB required some care to avoid damage. I hope you'll forgive all the technical language used here.

Getting the new chip back in was rather more difficult. Despite my slow and careful approach a single pin managed to get slightly bent where it was hidden behind an upstanding component. I removed the chip again, straightened the pin and reinserted. Careful examination of each pin through a magnifying glass followed as even more gentle lowering progressed..

With frequent grasps of a nearby radiator pipe, in case of static, I finally managed to get the chip safely back in. A final squeeze ensured it was as flat as it would likely to go without the assistance of a heavy workshop vice or G-cramp.

This is obviously the sort of task which gets easier with practice. This was my first ever attempt at chip insertion so one slightly bent pin was a slightly better score than the experienced chap on a YT how-to video.

Now all I have to do is light the blue touchpaper and wait for the magic smoke to rise.

To my shame I have all but forgotten the instructions for use of the AWR drives. After the initial excitement I have been too busy making castles in the air to do any observing.   

I did remember the sequence of ever-slower drive speeds for centering the moon in the eyepiece. So I do seem to have a few remaining marbles left.

The quality of the PCB and general construction is very satisfying to this relative electronics novice. Having seen the insides of quite a bit of hifi equipment and computers, as each eventually died over time, I'd call this AWR kit a very neat job. I wish my photography did it better justice but the light is poor as the sky darkens for a wet day with heavy, thundery showers.

I'll update this piece after testing for magic smoke.

Click on any image for an enlargement.
*

Building the Octagon Pt.64 Kung Fu wobbling.

*

I have delayed publication of two posts on building the roof until I can obtain some Cellotape. Paper models of bi-pitched octagonal roofs, using only copying paper, proved hard to work with in a Cellotape vacuum. You need tabs for gluing, as any child brought up on cardboard cereal packet models, knows full well. Moving swiftly on:

It was late afternoon when I spotted a pale half moon only 20 degrees up and almost due south. The earlier rain had gone with well spaced plates of thin cloud crossing from the SW. 

Contrast was bound to be low but it must be worth a try. So I dragged a large cardboard box of drive boxes, cables and eyepieces 'upstairs' and ran an extension lead.

The milky Moon didn't want to stay in the field of view for ten minutes so I needed to improve the haphazard mounting alignment. My laser rangefinder has an accurate clinometer so I started by getting the PA up to 55° instead of whatever it had drooped to since first being perched on its pyramidal stilts. I arranged the telescope over the mounting and pointing [roughly] at the pole to use that as a huge lever. Just one of the advantages of oversized ATM refractors. Don't do this at home without adult supervision! 

The tracking was slightly better after that but still wandering gently away. I had set up the azimuth to some arbitrary pointing at the solid canopy of trees. So out came my compass. Once I had distanced myself from the massive steel OTA [tube} I could safely assume the needle was pointing north. Then I had to turn the whole mounting so that it almost matched my arbitrarily aligned pyramid. Magnetic deviation? No idea until I look it up. I'd have to stay up until midnight for it to get dark enough to see the Pole Star and it was getting cloudier by the minute. There was also the matter of showering before dinner and could I possibly fit both in before July became August?

There followed further trials with the AWR paddle and several unexpected bleeps from crossing multiple meridians. The Moon now stayed "almost put" in the field of view of a 32mm Plossl at 69x. As I returned to the eyepiece after going back down to fetch a hex key I noticed the image was wobbling briefly. Only for about 2 seconds with rapid damping but it shouldn't be happening. I checked carefully around the floor's clearance gaps from the pier's pyramid posts. Weird, but everything looked fine.

I stamped my foot. Image wobble. I stamped again. More image wobbles. Then I notice my Bailey Bridge of 2x8s were almost invisibly shorting the ladder to the pier over the stairwell. I swapped one 2x8 for a 2x6 and moved them both away from the pier legs. Stamping my feet and even jumping up and down now showed no image movement. It's lucky I have almost total privacy in my back garden or the men in white coats would have been only a phone call away by now. "He's having another temper tantrum up on his tree platform!!" Little do they know..

Time to check the pier itself for stability. Being so far from the pier at the eyepiece on the end of a 7" f/12 refractor called for special measures. My head would almost have been touching the observatory wall had one existed. So I gently tapped the pier with the bottom of my shoe at full leg's length. The Moon wobbled by 1/8th of the field of view but rapidly settled. Hmm. I gave it a harder kick. Much the same.

No more infantile sissy kicking! Time for a serious bit of Kung Fu testing! I gave the pier a couple of really hefty Hollywood style, full-on, martial arts kicks. The same moon wobble occurred, but died back to completely still within two seconds. Can you kick your pier viciously with such positive results? How tall is your pier? I think we can safely say that mine gets the job done.

Early wind was slowly reducing as I played endlessly up in my open tree house. Earlier, 15mph gusts had caused a short wobble of the moon in the eyepiece. Again it was rapidly damped. If I touched the focuser I could instantly stop the wind wobbles. There was no problem holding the focuser continuously as far as the image stability went. I wouldn't be doing any imaging with such [slight] wobbling but it was still perfectly usable for visual observation. 

There is definitely some backlash at the worms because I never finished the screw pressure adjustment. The unwanted freedom at the wormwheel teeth may have allowed enough slack for the crosswind to affect the 7' long 8" Ø tube and added 10" dewshield.

I pushed the power up to 110x and the moon remained sharp but still pale and subject to rapid thermal 'boiling.' The Moon was by now sinking ever lower to the West, the nearby trees beckoned and the cloud was becoming a case of more often than not. I packed everything back into its cardboard box, put a bin bag on the vertical OTA and went in for a late shower and dinner. This is not really my idea of instant gratification. I want to be able to leave the cardboard boxes upstairs safely protected from the weather.

Click on any image for an enlargement. 
 
*

29.7.17

Building the Oxtagon Pt.63 Dimensions of roof panels.

*

Our octagonal roof has a nominal diameter of 10'.

If I choose a latitude of 45° for the roof "bend" I need to know the octagon's Circumcircle of that latitude. 

The small circle radius = RCosf . Where f is the latitude of the nominal [hemi]sphere containing the double pitched, rotating roof.

Cos 45°= 0.7071

If our octagon's Circumcircle is 10' then R = 5, then the small circle at 45° = 7.07' Ø.  [It doesn't matter where in the calculation you multiply or divide by 2 as long as you remember to.] 

This small circle marks the points of the octagon. The distance between these points is 7.7' x Pi  / 8. The circumference divided by the number of sides. 22.21' /8 = 2.77'. This is the width of the base of the eight triangles forming the upper pitch of the roof. Or the top width of the trapeziums forming the lower pitch of the roof.

The width between the points of the equatorial circumcircle is obviously Pi x D /8.

3.142 x 10' /8 = 3.9'. This is the width of the bottom of the trapeziums. 

The height of both the triangular and trapezoid roof panels is 1/8 of the circumference of the circumcircle. So again = 3.9'. We now have the nominal sizes of the roof panels.

____________________________________

When I first tried to model my intended roof, to check for the desired overall shape, I was assuming that panels would be straight-sided isosceles triangles. I obviously wasn't thinking straight. Straight sides would only be the case if they formed a straight-tapered roof.

Whereas, bi-pitched roof panels are bent at the junction between the upper triangle and the lower trapezium. The width of this joint must be greater to allow for the depth of the bent form. [Spherical geometry?] This width should, of  course, be 2.77 units. Exactly as calculated above. Fingers firmly crossed as I take scissors to cereal packet cardboard in the time-honoured way.

Whoopee! Almost unbelievably, the model appeared exactly as I was expecting! Next time I'll add double tabs to my four sided units so that I can use clothes pegs on the triangle's inside joints. I'm waiting for the PVA glue to dry on three triangular panel tabs so I can finish the model and take some photos as proof of my calculations.

Click on any image for an enlargement.
*

Building the Octagon Pt.62 Octagonal rainproofing.

*

An octagonal roof of nominal 10' "diameter" [Circum]circle.

Pi x D = Circumference. 

3.142 x 10' = 31.42'.

31.42' / 8 = 3.9275'

3.9' < 4' = so can be obtained from a normal [4'x8'] sheet  of plywood. A fortunate coincidence. Thin, 4mm [5/32"] Birch plywood is also available in 5'x5' sheets. 1.5m x 1.5m.

A rotating octagon scribes two circles at its "circumference." A smaller one [Incircle] forms Tangents at all the flat sides. While a larger circle [Circumcircle] contains the Points of the angles between the eight sides. This is important because a rotating octagonal roof has a series of [longer] overlapping points and shorter radius tangents. As does the building supporting it.

The octagonal roof's Tangent Incircle must always be larger than the supporting octagonal building's Point Circumcircle. So that rain is always shed outside the building. If the roof's Tangent circle is the same size as the building's, then the roof Tangents will fall inside the building's Point circle at 8 successive angles of rotation. Buy 8 buckets to catch the rain? Or make the roof suitably larger to avoid problems. [P&T are my capitals added here for clarity.]

The radius of the [Incircle] is found by multiplying the Circumcircle by = 0.924.
So a 10' diameter octagon will have a 9.24' diameter Incircle. A 9" difference or 4.5" difference in radius. That 4.5" difference is crucial to making the rain fall outside an octagonal building from an octagonal roof. Some extra overlap is obviously desirable to ensure clearance for flashing attached to the edges [tangents] of the roof where the building's Points pass under it during rotation.

Click on any image for an enlargement.
*

28.7.17

Building the Octagon Pt.61. Put two wheels on your hatch?

*

A CN forum member suggested inserting a pulley into the hatch lifting system. At first I could not see the advantage. The pulley would halve the force applied. The real trick was in grabbing the pulley idea and running with it. Am I allowed to say that at my age?

Why have bell-cranks when you can have two pulleys? Imagine a large, wooden pulley attached firmly to the hatch on each end of its hinge line. Now wrap two ropes around and anchor them to their own pulleys. Using only one pulley would tend to twist the hatch and cause all sorts of diagonal force problems. 

Two pulleys make a balanced system provided the wights are fairly equal. But it gets better! Now hang one rope loop over both pulleys and anchor each end to its own pulley. At the bottom of your loop you hang the hatch's lifting weight. The pulley and single hanging weight provide perfect balance between the two pulleys. All asymmetry of forces is vanquished.

Why have pulleys on the hatch instead of simple bell-cranks? Because the limited range of movement of the bell-cranks was causing considerable changes in lifting force as the hatch opened. Pulleys provide constant force because the lifting moment never changes. Moment = Mass x pulley Radius. Both remain unchanged.

The lift begins from fully closed and continues steadily throughout the hatch's total movement to fully open. 

Obviously the pulleys must be strongly attach to the heavy hatch. The ropes are easily anchored to wooden pulleys. You just drill down from the grooved rim to exit out the side. Or drill down straight into a pre-drilled and slightly larger cross hole. You push the end of the rope down the rim hole. PUll it out the side and tie your knot. Now you pull back on the rope at the rim and your rope is safely anchored. The knot can never escape unless you take the weight off the rope and pull it out of its cross hole.

The image shows a quadrant arrangement to help to lift the hatch. The quadrants start high and descend into slots between the boards of the floor behind the ladder. So the slots are never open. Each end of a cable or rope is anchored in each quadrant. The weight hangs from a single pulley to balance the forces on each quadrant and thus the hatch.  

The size of the pulley wants to be the largest possible which won't intrude beyond the handrails.  Why turn wooden pulleys when you have plastic or steel wheels with the tyres removed? Three laminations of plywood with a smaller central disk automatically makes a pulley. 

Time to start measuring how large a pair of pulleys I can have. But why make do with complete disks when you can have a larger quadrant? One which projects backwards from the ladder and out of the way? The quadrant just wants to have the same angle as the total hatch movement =120°.
Constant force x Constant radius = Smaller weight = lower wear.

*

27.7.17

Building the Octagon Pt.60 A new hatch not plotting my downfall?

*

Thanks to the ever helpful souls on the CN Observatory forum I have overcome my "Sidewinder" larch, hatch inertia. The sideways opening hatch always had problems. Not least its weight! 

A forum member suggested a forward opening hatch worked well for him so I had another try myself. 

I needed a deep foothold at the top tread. I also needed the hatch to be flush with the obs. floor level when closed. 

The answer was to pull the ladder back by 10cm [4"] and lower it by 1" to bring the top tread down by the obs. floor thickness. By fixing a 2"x4" beyond the ladder top it now had something really solid to rest against. The handrails had flexed somewhat when the ladder was in use. The 2x4 also provided my vital deeper landing at top tread height.

The 1" difference from obs. floor level didn't seem to matter, in practice. It was much easier to use the top step going both up and down the ladder. Previously I would step up from the next tread down and move bodily sideways and upwards at the same time. Even with lots of practice it never felt very natural. Particularly as I was also avoiding the handrail with my feet at floor level. Now I can comfortably use the top tread for the lateral transition to obs. floor from much closer to the same level. There is no simultaneous, upward body movement required.

The new hatch can now be laid on the top tread [when closed] without it becoming a nuisance or danger when open. It will be hinged at the edge of the existing larch boards as shown. The mock-up in the image uses a smaller scrap of 18mm ply. In practice the new hatch will be the full width between the handrails at 50cm wide. Though there is room for 53cm the hatch might strike the handrails as it rises and falls.

The new hatch can also be made much longer so that it fits inside the pier's "kennel opening." This will provide a much greater level of safety when the hatch is closed. With a greatly reduced danger of my falling through the opening. A foot bumper rail is now easily possible without it lifting the hatch off the floor when open.

When opened, the new hatch will lean safely against the inside of the observatory wall between and beyond the ladder handrail extensions. Where it can be latched for extra security if so desired. I get to keep my full height handrails and close off more obs. floor gaps in one go. Now I just need a much lighter hatch material than the heavy larch floorboards. The hatch can be braced with stiffening strips on the hatch underside where it clears the top ladder tread. The foot bumper will provide lateral stiffness.

I have some 12mm [½"] plywood to make a trial hatch. If I shorten one length to clear the ladder's top step I can have 3 layers for 1.5". Perhaps it doesn't need to be that thick? The weight is a serious issue even if it is easier to lift it forwards. Since the ladder will support the hatch right up to the change in thickness two layers of plywood might do.  The larch was only just over an inch thick but weighed a ton. With noticeably wide variations between samples.

There is the possibility of using spring hinges, gas struts or counterweights to aid lifting the new hatch. Since it the angle of rotation [around the hinges] from fully closed[horizontal] to fully open exceeds 90° simple levers and attached weights won't help. Ropes might get in the way when the hatch is closed.

Weights or levers could pass through the unused floor area beyond the top of the ladder. Either through slots or holes for rope.  I'll have to give this some thought. Levers attached to the hatch with secondary weight levers passing through slots might work.

Here I have shown the hatch rotating around the hinges with fixed [primary] lever[s] attached at an angle of 135°. Ropes or a second set of freely pivoted levers carry counterweights to almost balance the hatch. The primary levers must be able to pass through slots in the floor beyond the hinges or they will strike the floor. Easily arranged simply by separating the larch boards as required.

Black is hatch closed. Red is hatch lifting. Green is hatch fully open and parked against the obs.wall.

Note how the balancing force increases rapidly as the hatch starts to rise. Then steadily falls again as the hatch nears the vertical. [Where least effort is required.] Since the [counter]weight remains constant only the moment changes.  [Mass x distance from the pivot.] The moment of force is measured as the horizontal distance between the hinge and the perpendicular of the hanging weight.

The angle between hatch and fixed primary lever[s] can be adjusted to maximize the lifting force from first opening. The limitation being the primary lever[s] becoming vertical in a downward position and providing zero balancing force on the hatch. Perhaps even blocking it from further movement. The advantage of this simple system is that it is foolproof and easily adjustable. There is nothing to go wrong provided the rope doesn't break. The weight, on its rope, can hang inside the observatory wall, under the ladder, well out of the way. If it is arranged to hang near the floor the weight can rest on the ground at its lowest point. Taking a well earned rest when it isn't needed to balance the weight of the hatch. 😉

I feel much happier about the safety and comfort of the new hatch arrangement. While I could lift and lower the old hatch it was an unnatural movement, sideways at arms length. Particularly if I was not to simply drop the hatch onto the obs. floor with a loud crash. Hardly sensible, when it could be happening in the middle of the night!

My search for a 135° angle bracket was fruitless. I scoured two large, DIY outlets in vain. Ideally I would find 1" wide strips, like a normal 90° corner brace, but with a 135° angle instead. These could be screwed to the outside edges of the hatch to project at 45° upwards and away from the hatch hinges. Steel would be best for stiffness under load but aluminium might work if thick enough to resist twisting sideways. Arm lengths of 10"-11" would be ideal.

Sudden screech of brakes!! A suggestion from another member of CN forums brings a new an potentially superior method of lifting the hatch. See the next exciting episode: Pt.61? Really?

 
Click on any image for an enlargement.
*

Building the Octagon Pt.59 Painting in the details.

*

By means of cloning I can paint in surface texture "borrowed" from other parts of the same image. In this case the weathered surface of 15 year old plywood cladding boards, with machined grooves at 4" intervals. 

I tried other surface treatments and colours but much prefer this one. The veranda "railings" were added using a simple line drawing tool. As was the border to the door.

Painting a snow white dome on top of a freshly boarded observatory building was absolutely hideous! So the rotating roof needs much more gentle treatment in this rural environment. Unfortunately white is the best colour to reflect solar heat. Black is the probably the worst but many colours have odd absorption behaviour in the Infra Red.

The brand new plywood cladding boards can be given the mineral powder treatment. Hopefully to darken the surface to a weathered finish. Like the shed, but without the very long wait.

Perhaps even the same boards could be used for the roof but sloping to shed rain. While still providing internal clearance for the telescope to swing.

An octagonal roof would need some overhang to ensure no rain can "run down the observatory's collar." An octagon has a smaller radius at the center of its straight edges but overhangs considerably more at the "points." There will be a series of eight rotation points where the relative points and straight edges overlap with large or small overhangs.
A double pitched octagonal roof can contain the radius of the swinging dewshield with reduced use of materials and volume compared with a simple taper. A dome can be thought of as an infinite series of flat surfaces too small to see individually.

I like the idea of an octagonal, double pitched roof for another reason. The panels which form its upper triangles and lower trapeziums are of manageable proportions in a 10' nominal diameter. The roof can thus be safely clad by a person working alone. Assuming, of course, still conditions are chosen to perform the task.

Here I have modified the roof of a gorgeous, old, wooden barn to an equal length, double pitched octagon. How can I simulate small "tiles" or "shingles" without adding massively to the roof's rotating weight or badly increasing solar heat gain? 

Click on any image for an enlargement.
*

26.7.17

Building the Octagon Pt.58 It's huge!

*
Wednesday: After yesterday's wet weather it was pleasant to be back into sunshine and light winds. You can keep the warm.

It occurred to me that I had no real feeling for the size of the observatory yet. The open sides were far too transparent to be able to pause the eye. One's focus passes straight through to the surrounding scenery. So I hung a new 3x5m lightweight tarpaulin where the obs. walls should be. These are about 4'6" or 1.4 meters high.

Remember that no roof or dome has been placed on top yet. This is just to simulate the observatory walls. The octagon is a little over 3m or 10' across. I don't think the image really captures the sheer scale of it. Perhaps I need more background? Or I just need to move further away with the camera?

This high level shot, from a dormer window, makes it look far more squat and heavy. That wasn't expected at all. I thought it would look far too 'lanky' on stilts. Though the dome/roof will make it taller again it should remain in balance once all the walls are covered.

My wife and I have been searching countless images online for inspiration for a more traditional roof style. A conventional, rib and gore dome is still possible but alternatives beckon. 4mm BWP Birch plywood looks like the best bet if I go for all flat surfaces. Some surface texture would be nice. Thin plywood 'fish scale' tiles perhaps? Though it would certainly add to the weight. Just another idea discussed in passing. Thatch anybody? I kid you not. Its pros and cons were discussed.

I had better get on with the veranda or it will never be finished! Yet another veranda side [6th] went far more quickly than previous ones. I shall be getting good at this just as I finish the last side. The latter is fast becoming a distinct reality. I just need to trim the right side of the last [7th] side and I have [almost] finished the floor boarding. There are a couple of short, diagonal joists to be fixed underneath and some gaps to be filled under the obs. walls. This is deliberate and depends on the space required for flashing. Only then can I judge the desired width of these narrower board 'fillers.' 

I have to admit that while the flooring looks okay at a glance this has been my least favourite part of the entire construction. You would not believe the number of times I have been up and down those 10 steps! The miter saw is downstairs in the shed next door and the work to be fitted and measured upstairs. With often only a hair's breadth [repeatedly] shaved off here and there it has sorely tried my patience!

Each 1"x 5" floorboard was usually long enough and therefore heavy enough to be a nuisance to carry up and down. Not to mention threading them all through the small opening upstairs going both ways.

The weather has not been particularly helpful. With quite unnecessary rain and equally unnecessary heat regularly changing places in the forecast. A wise man would [probably] have put a roof on it first. To afford shelter from both rain and heat. My practice on the ladder has made me confident enough to become careless. While my thighs have grown muscles many cyclists would envy! Am I having fun yet? 😎

Perhaps I should add that the upper braces on the obs. wall posts are for strength. I haven't cut the posts level with each other yet. Unfortunately it is too early to decide on wall height so I can't clad the obs. walls with plywood either. Think of the build as a journey of discovery. Not so much a collection of whims but something to be made up as I went along.


Click on any image for an enlargement.
*

24.7.17

Building the Octagon Pt.57 Plotting a hatch.

*

With all sorts of weird ideas about the necessity for 180° hinges floating about in my mind, I cycled the 10 hilly miles to a DIY outlet. Where I found some nicely sturdy hinges of ample thickness, width and girth and then rode home again.

It was only on arriving home that I discovered my lack of CSK head machine screws of suitable length and diameter. Otherwise I had made a sound choice in the hinges. Having clamped a board to the hinge and the hinge to the side board, I was delighted with the action. Which laid the [simulated] open hatch completely flat on the bare obs. floor to the south without actually hitting anything. Thin rubber pads might be useful if air pressure alone does not cushion the fall.

I proceeded to hack out a rebate to bring the first hinge flush with the board running alongside the stairwell. Round head machine screws were then used to temporarily hold the hinge in place.

Four 77cm [30"] lengths of near 12cm wide, larch floorboard were slightly too wide to fill the 42cm deep gap. Which allowed me to experiment with a 'tongue' projecting between the ladder handrails. I had decided to run the boards across the gap at right angles to the obs. floor boarding. My fuzzy, engineering logic suggested that four full boards would be stiffer than only two straps holding four short boards together. As would have occurred had I followed the E-W line of the rest of the obs. floor boarding.

Making the hatch/hole any narrower would have prevented my hands from safely following the handrails of the 60cm [24"] wide ladder. 77cm is quite an unsupported span but the larch floorboards are of ample thickness at 27.5mm. [1.1"] I shall proceed to complete the hatch and fit the second of the three hinges purchased. The third hinge may prove useful if two seem too flexible in practice.

Gluing the boards edge to edge and then gluing and screwing the two straps seemed like a good idea for extra strength. The hatch has been left clamped up overnight under a tarpaulin in case it rains. [It did!]

I shall, of course underlay each side of the stairwell gap with a solid ledge for the entire width of all four hatch boards to rest on safely and securely. I would not want to trust the bare hinges alone without some underlying support. A couple of abbreviated joists might be useful nearer the pier legs.

Tuesday: I was just trimming the ends of the hatch boards with the router when it started raining hard. Earlier I had centered the ladder in the obs. floor cut-out as it seemed to have drifted 3cm to the left. The hatch is now pretty weighty to carry 'upstairs' with nearly 14' of 1"x 5" larch boards all stuck together!

The lower images show the hatch trimmed and the hinges temporarily screwed in place. The hatch is remarkably easy to manage with the hinges taking most of the weight and alignment duties. It certainly drops with a bang when released from vertical but that can be overcome. It feels very solid when I walk back and forth. I temporarily pushed a couple of lengths of 2x6s in near the pier posts as extra support joists. Just to spread the load between more of the planks while maintaining isolation clearance.

Once the hinges are flush only the slight bulge around the hing pin will protrude from the obs. floor. Being low and well rounded it should not present a serious trip hazard. Particularly once the hinges are sunk to their correct depth. [Flush.]

The overall dimensions of the hatch ended up being quite modest. I had imagined there was a far larger hole in the obs. floor. Notice how little trimming was required to get the board nearest the ladder to close off some of the gap while retaining a full width of floor board. I used the jigsaw, with a 30° angle on the base plate, to undercut the notches to clear the ladder's handrails. Thankfully, my fears that a large and very heavy hatch would need mechanical assistance to open and close were completely unfounded.

Work has been interrupted at irregular intervals by the rain today.

Click on any image for an enlargement.
*

23.7.17

Building the Octagon Pt.56 Eastern or Southern access ladder?

*

With the through-the-veranda ladder idea sinking fast, I looked at the opposite way of doing things. Until I bought the ladder I had assumed I could never use it for access through the workshop/shed. However, with the steep rise of 70° the ladder covers only 1.5m [5'] along the ground for its 3.5m total rise. [To the top of the handrails.] 

If the obs. octagon is 50cm from the shed then the ladder will consume only 1m of shed width inside. This can be catered for with an internal re-arrangement of my hobby life's detritus. 

The image has been adjusted for perspective to put the ladder at the correct distance from the viewer. The exact lateral position of the ladder will depend on the size of the "landing" at obs. floor level. There is no veranda on this side.

The most difficult part of this, by far, is making the shed roof waterproof despite the ladder piercing it at eaves level. This matter should not be underestimated or the shed would quickly become unusable. A simple flap wont do because the ladder passes trough the hole in the roof. So the ladder would need a fully enclosing 'box' sealed to the shed roof with flashing, etc. The large triangle filling the open space between shed roof and obs. would almost certainly be an awful eyesore! A glass sided arrangement is not possible on the grounds of cost alone. However tempting, this is not remotely an easy option!

It might be possible to have the ladder rising from just inside the shed. This brings the ladder very close to the pier at the top. Weatherproofing of the shed wall is far more easily arranged with a simple porch between the two buildings. However, this idea needs a very large hole in the obs. floor for the user to climb through. With the joists cut away, serious thought would need to be given to the structural strength of the building. Reinforcement would be inevitable and difficult to arrange.

Another alternative is to have the ladder exposed in the foreground and approaching the building from the south. An access door would then be required in the SE face of the obs. octagon wall. This would leave the ladder exposed to the elements with all that entails. The SE veranda becomes the [potentially slippery?] landing for the ladder.

I have drawn in the ladder to scale to show the general arrangement. Any attempt to roof over the ladder for protection is almost bound to be visually, rather clumsy. Advantages include easy and direct access. Disadvantages include exposure to icy weather/snow. The ladder could be removable for security. Or a plank padlocked over its face to avoid uninvited guests on the first floor. This arrangement feels clumsy just looking at the image above. From an oblique angle it would look better thanks to the relative sophistication of the ladder.

Yet a fourth idea would be to partially hide the ladder in the gap between the buildings. This would require modification to both shed and Obs. due to the slightly greater width of the ladder. Arranging guttering might be a problem.

Unless I can think of something else, it seems the present, internal ladder arrangement is fixed. It is well protected from the winter weather and the building is secured by a solid, ground floor door.

Despite the sometimes thundery showers I was able to fix some floorboards with ss. decking screws. As well as blogging my endless thoughts on ladder layout, obs. access, and taking loads of photos, of course.

Now I have to seriously think about a rotating roof or dome. The pier is slightly higher than planned so a 3m semicircle is struggling to contain the dewshield. See image. I have marked head height at 1.8m with a flower pot on a pole on the right.


Click on any image for an enlargement.
*

Building the Octagon Pt.55 External stairs?

*

While thinking about the hatch over the stairwell it seems I was compensating for poor siting of the ladder. I had hacked away at the pyramidal pier cladding just to allow me to climb through it. The safety hatch is only there to stop me falling down the open stairwell during normal use of the observatory! How foolish is that? 

The image shows how the ladder could be arranged to climb up through the western veranda. The veranda and ladder width match each other nicely.

With the ladder piercing the veranda, the user would pass between the handrails as they reached the veranda level. The handrails would offer a natural stop to anyone approaching the ladder [and potential floor opening] along the veranda from the far side. 

The ladder can be protected from the elements with a covering wall/box of grooved plywood to match the rest of the building. 

Access to the observatory proper inevitably demands the user ducks under the dome ring. Without modification, the present, maximum clear door opening height would be a miserable 1.3m. Which is only 4'3" in Old Money. The top of the ladder would also need weather protection to avoid snow and ice accumulation.

The Head Gardener doesn't like an external stair for umpteen, very sensible reasons.
Quote: "Stop being a girl!" Which made me laugh. 😊 She is usually so protective of my physical welfare, but it seems aesthetics takes precedent this time.

Click on any image for an enlargement.   
*



*

22.7.17

Building the Octagon Pt.54 Stairwell closure hatch?

*

I think I may go with a side hinged trapdoor to cover the large hole in the obs.floor over the stairs. I tried hard to visualize one hinged beyond the ladder top but it just didn't work. It was impossible to match levels without it being above or below the floor surface or matching top step of the ladder. A side hinge can be arranged to pivot exactly at floor level. That is the only way I can imagine the finished trapdoor remaining at surface floor level when closed.

This image, looking north [original taken in the wet] is a mock-up of how the trapdoor would close off the open stair well. This will allow safe passage across it in the dark. The ladder now needs to be raised 50mm [ 2"] to bring the top tread flush with obs. floor level. Vibration-isolation clearance is required for the pier on the right.

Because the ladder is sloping [at 70°] it might be possible to extend the trapdoor to the left to reach between the handrails. I'd have to check the geometry to see if it is possible. 180 degree [flush] hinges would be required to allow the hatch to lie flat on the floor towards the south when opened. [i.e. Nearer the camera.]

I don't want a small step or drop to catch me unawares in the dark when the hatch is closed. Hatch closure should become routine during preparation for observing or imaging. A light/LED[s] in the lower room will ensure the hatch does not remain open after dark without the large hole in the floor being very obvious. A micro-switch to extinguish the light/LED[s] on hatch closure is not difficult to arrange as increased insurance against free-fall skydiving.

It might also be possible to extend the hatch slightly inside the pier cut-out. A simple safety bar could be arranged to pivot across the pier "kennel" opening when the observatory is in use. A suitable slot would allow the bar to rest on a screw when in the lowered/closed/horizontal safety bar position. A door on the pier cutout is also possible. The sloping surface would ensure it stayed closed without needing a latch or lock.

This is the original image with the stairwell open after rain. The large hole is an obvious danger in the dark. I left the jutting joists overlong in case I needed their support for some reason.

The floor hatch is likely to be quite weighty if made of larch boards. So a flush ring-pull handle [or knotted rope?] would be useful for the initial lift. A rope could be clipped into an open receptacle ring on the pier without compromising isolation. That would make it easy to control the hatch without stooping over the gaping hole in the floor.

There are presently gaps outboard of the ladder necessary for safe use of the handrails. The previous hole, between the top of the ladder and western wall, is now covered in floor boards.[Visible left beyond the top of the ladder.] There will probably still be a small open gap between the ladder handrails at floor level but it will be all but inaccessible to clumsy feet in the dark because of the jutting handrails themselves. My main worry is that I might be concentrating on the view in the eyepiece as I move across this danger area in the dark. Without some means of semi-automatic protection I am in increasingly likely to do a header if not protected from myself with increasing age.

Click on any image for an enlargement.
*

Building the Octagon Pt.53 A plea for empathy.

 *

Friday: Another side of the veranda was boarded before lunch. I  have discovered a superior way of fitting the boards: Ensure the miters on the left match the previous veranda side. Then leave the other ends long as each board is spaced apart with nails. Once satisfied with positions, spacing and fit I draw a line from outer corner to the edge of the octagon post across all five boards. Then I take the boards down and cut to the line on the miter saw. Hey, presto!

It takes a fraction of the time compared with cutting each board individually to length and miter angle before moving onto the next. The method was suggested by CN forum members but it didn't really sink in as pertaining to my particular problem.

I couldn't run a circular saw across because of the posts and rim joists. I tried using a jigsaw but the joists got in the way and the mitered edges were always wavy. Remember that I am not a trained carpenter working day in and day out. I am just a hobbyist with a few tools and limited knowledge and skills and working occasionally on jobs often years apart.

In the afternoon I laid the 6th side of the veranda floorboards. I still need to add another diagonal joist to catch the overhanging ends of the boards. I had to cut the miters on the left several times before I was reasonably satisfied. Each time that shortened the boards. Each time I had to replace all the loose nails I was using as spacers. If the first board is not parallel with the rim joist then the miter angles change every time. And it shows! I was even screwing all the boards down before marking the miters on the right hand end with a straight edge. Just to avoid making careless mistakes in angle and length.

The DeWalt miter saw can cut perfectly straight and is adjustable to 1/2 a degree. It can shave off a couple of thousandths of an inch if needed and often is. Despite the clever projection of the blade line onto the work to be cut, it still takes practice and more practice, to get accurate cuts precisely where required. The fences are short and the planks are often long and more often than not badly bowed. Choosing which side to use is often compounded by grain, splits, cosmetic damage or loose knots. Ideally all bowed lengths should be 'nested' to avoid closed or open gaps in the middle or ends. All this may be second nature to an experienced woodworker but is not necessarily true of a hobbyists working on an occasional project.

Click on any image for an enlargement.
*

20.7.17

Building the Octagon Pt.52 Veranda progress.

*

With the threat of rain hanging over me I pottered on trimming back the beams where they met the shed eaves. That resulted in further work on the 'corners' [angles] nearest the shed.  I need to make room for gutters on the shed. It was never necessary in the past but I don't want rainwater splashing up or lying in the gap between the buildings.

I also added a floorboard at the front where there was a large gap between obs. floor and the veranda. The flashing will cover the boundary between the face of the obs. and the veranda. The flashing will be tucked under the nearest veranda floorboard to ensure rain running down the obs. wall is carried away. Hopefully avoiding any rain running back into the observatory itself. 


Earlier on, I was afraid the bulky 2"x8" front beam was going to dominate the view of the observatory. However it now lies in deep shade under the veranda floor. So it will be lost in the front wall of the lower building. I have no experience to guide me. So there can often be ten trips up and down the sturdy aluminium ladder.

Boarding the veranda has been a slow and difficult process. Pride demands reasonable joints at the angles. As did the dominating miters at the change of angle on the rim joists. The weather and natural changes in the timbers has altered the accuracy of my patient cutting.

Crucially, I never used a cord line to mark the perimeter of the octagon posts at the beginning. So I never really had an accurate guide to the external dimensions of the veranda. I just made them look right to my own eyes. Looking above and below the veranda beams at the octagon beams allowed me to see non-parallelism if it existed. I had already checked levels and uprightness of every post, joist and beam using reasonable quality spirit levels.

In practice the perimeter of the veranda is not truly parallel nor even truly concentric to the octagon sides. I don't think this really matters because the wedge shaped boards nearest the octagon will be concealed by the flashing.  I don't think it would serve any useful purpose to fix the veranda to true geometric precision with the octagon. It would probably just look twisted and perhaps a bit mean.

I would certainly recommend that anyone building such a structure set out the project properly with cord lines. Their absence in my case was simply ignorance of normal building practice. The concrete footings were never precision placed but could have been by using cord and posts. I tried to hold the building accurate by measuring between posts with a tape measure. The braces between posts were mostly [almost] the same length. None of this affects the strength nor appearance of the structure. Which is/are the important thing[s].

In my own defense I was lax at times because the support posts were bent or twisted or became twisted in use. Where do you measure from when the start and end points have rotated or moved?

The Head Gardener makes a regular inspection of progress and workmanship.

Those who work with wood every day would know whether to reject such material at source. Or know how to deal with natural changes in the materials. Or how not to build using such simplistic methods as my own. Scaffolding would have saved me not being able to reach the upper floor until much later. Simply because I was unable to lean a ladder up against such a loose and flexible structure.

Click on any image for an enlargement.
*

15.7.17

Building the Octagon Pt.51 Fixing the Obs. larch flooring.

 *

Saturday 15th: In warm, 72F sunshine, I am finally fixing down the obs. area, larch floorboards. I am using wedges against the octagon support posts. This is to push all of the boards tightly towards the center to close up all the gaps. That meant trimming larger notches around the pyramid pier legs and re-cutting the end miters.

View from the shed end gable nicely shows the construction of the raised platform. Flooring the veranda comes next following the recent 'eco' treatment.

I am using 3" stainless steel, decking screws, with small Torx heads. So that [in theory] they sink neatly out of sight below the surface. It makes a pleasant change not to hear creaks and rattles as I walk about up there. It may even stiffen up the building in a belt and braces sort of way. You would not believe how quickly floorboards eat up a whole box [and a half more] of 250 screws!

Half an acre of larch boarding.Well, 3m or 10' across. You do the maths.

Now I have to go back around with the DeWalt on "drill" setting to pull many of the screw heads down. Lots didn't go into the surface even on a "15.5" maximum torque, driver setting.

I love the way the larch boards have darkened the grain pattern with exposure to UV sunlight and rain. A brand new look was never in my list of desires and modern "knotty pine" boards can never look the part. I want the floor to look "antique" and "lived in" before I even start using the observatory in earnest.

Sunday 16th.  A wet morning is forecast. Persevering with floor boarding.

Monday 17th. Two sides of the veranda now boarded.

Tuesday 18th. Veranda boarding sneaking towards completing a third octagon side. It needed joist height adjustment to cater for half inch upward bowing!

Thursday: Another view from the shed end gable showing three sides of the veranda almost complete.

Unsupported board ends probably need a new, short, diagonal, 2x6 joist. The far end coincides with an expected, railing corner post so I am undecided as to what to do next.

I have painted in the lines of the existing parallel joists in red. The run of board, end miters lies at an angle between the two joists. [Green] Even a new 2x6 joist [blue] in the middle, between them, isn't really a perfect solution.

Though it catches most of the longer overhangs. Perhaps I should use a diagonal 3x6 or even a doubled 2x6 instead? That can be fixed to the corner post later with a matching socket. The inside end can be mitered and screwed to the nearest joist beside the octagon post.

I went with the green option but had to go out to get more 2x6. The image alongside shows how I caught most of the overhanging boards with a mitered, diagonal 2x6. It is a compromise but the overhangs are all quite modest. I shall notch the outboard end into the railing post.

I don't want to start installing the corner, railing posts until I have the roof/dome on. Otherwise they are bound to get in the way of lifting the roof sections. Probably by sliding the panels up pairs of spaced ladders.

Click on any image for an enlargement.
*