Roman Quadrireme Galley by Ian_Grant - 1/32 Scale - RADIO

[Ian_Grant]

Richard,  thank you so much for the detailed data on the stroke! Fascinating to study. I can see I have a lot of coding to do especially in details of the lift motion. I just started building a jig with a reme of 20 oars; plan is to use some sort of  strain gauge to measure sweep force required to move blades in water. This equates to rowing with no response from the ship ie no acceleration, which will give me the maximum possible force. I've been looking at the Hitec HS-755MG and HS-805MG giant analog servos, hoping the 755 will serve because the 805 draws half an amp just to move its arm with no load (!). But we'll see what falls out of this next test. I'd rather have a big motor loafing along than a smaller motor labouring, provided the big motor isn't too power hungry.

 

As for the mechanism, I drew up some ideas using servocity structure and motion components. The vertical motion will definitely be linear bearings on 1/2" vertical shafts. My prototype oar attachments were crude, just a flathead screw left a little loose so the oar could tilt. I now plan to mount the oars in a u-channel which allows sweep movement only; the u-channel itself would have to pivot along its axis to allow the lift movement. I considered mounting shafts axially at the ends of the u-channel, said shafts passing through bearings mounted in pillow blocks, but that's a little pricey. The fallback is to simply mount the u-channel on hinges so it can tilt around as the beam moves vertically.

 

Again, thanks for the useful data. I'll need to improve the code defining the stroke shape, for sure.

 

By the way, there is a GPS "shield" for Arduino. Hmmmm.......

 

https://www.robotshop.com/ca/en/sim28-arduino-gps-shield.html?gclid=EAIaIQobChMIlpWOlvS29wIVRCCtBh18uAqKEAQYAyABEgK3LPD_BwE

 

[Richard Braithwaite]

7 hours ago, Ian_Grant said:

Richard,  thank you so much for the detailed data on the stroke! Fascinating to study. I can see I have a lot of coding to do especially in details of the lift motion. I just started building a jig with a reme of 20 oars; plan is to use some sort of  strain gauge to measure sweep force required to move blades in water. This equates to rowing with no response from the ship ie no acceleration, which will give me the maximum possible force. I've been looking at the Hitec HS-755MG and HS-805MG giant analog servos, hoping the 755 will serve because the 805 draws half an amp just to move its arm with no load (!). But we'll see what falls out of this next test. I'd rather have a big motor loafing along than a smaller motor labouring, provided the big motor isn't too power hungry.

 

As for the mechanism, I drew up some ideas using servocity structure and motion components. The vertical motion will definitely be linear bearings on 1/2" vertical shafts. My prototype oar attachments were crude, just a flathead screw left a little loose so the oar could tilt. I now plan to mount the oars in a u-channel which allows sweep movement only; the u-channel itself would have to pivot along its axis to allow the lift movement. I considered mounting shafts axially at the ends of the u-channel, said shafts passing through bearings mounted in pillow blocks, but that's a little pricey. The fallback is to simply mount the u-channel on hinges so it can tilt around as the beam moves vertically.

 

Again, thanks for the useful data. I'll need to improve the code defining the stroke shape, for sure.

 

By the way, there is a GPS "shield" for Arduino. Hmmmm.......

 

https://www.robotshop.com/ca/en/sim28-arduino-gps-shield.html?gclid=EAIaIQobChMIlpWOlvS29wIVRCCtBh18uAqKEAQYAyABEgK3LPD_BwE

 

You can get strain guages for full size oars and I have thought about making a full size mockup of a single rowing position to measure forces to validate my computer simulation. However, it is probably impractical to measure the strain in a model oar shaft as the loads on the oars are really tiny.

For example the maximum thrust I measured in a zero speed trial of my Mk 1 galley (750mm long, 500 grams in weight, 12 oars, 150mm outboard length at 45 strokes per minute) was 0.0123N (measured using a fine thread attached to a post on the stern running to a pully with a hanging weight). So that's only 0.001N/oar!

The average speed achieved by this galley at this stroke rate was 0.09 m/s.

If we say that this galley was at a scale of 1:24 then this speed equates to around 2.16m/s or around 4 knots (if we scale speed by length). This isn't particularly fast, but then a constant speed circular oar motion is not particularly efficient!

The motor in my Mk 1 galley was way over the power required, but that's not really a problem (I wasn't interested in efficiency here) but it does ensure that any friction in the mechanism (much more significant than the propulsive forces) is easily overcome.

(Now if you fitted your rowing machine to a full size galley, or a really large scale model, the loads would be larger and easier to measure!!)

It might be better to measure the boat speed somehow and program your oar motion accordingly rather than try to measure these tiny forces?

Alternatively you could increase the  speed of a small scale model, but that would require ridiculously high stroke rates!

 

Ive Just found a simple spreadsheet model that I used to predict the oar forces in the first trial (zero speed 45 spm) I referred to above:

 

RowingMk2Trial1.xls

 

And here is the spreadsheet set up for the second trial (steady state speed at 45 spm)

 

RowingMk2Trial2.xls

 

Since this rowing machine moves the oars at a constant (circular) speed no matter what the boat is doing its relatively easy to simulate with a spreadsheet. Still I was quite pleased at how close my spreadsheet came to the measured performance of my little model. Adding a human being into the simulation does, however, makes it a lot more complicated!!

 

 

 

Edited April 28, 2022 by Richard Braithwaite

[Richard Braithwaite]

This document contains a bit more detail on the trails and calculations described above and application to a working model of Olympias using the elliptical machine shown in the video on a previous page of this thread.

 

Rowing Machine Calculation.pdf

 

One interesting finding is that with The top tier only (i.e. 62 oarsmen) the average speed is predicted at 6.53 knots, 

tis increases to  7.72 knots with all 170 oarsmen. So a significant increase in speed, but not as much as one might expect for all these additional oars. The main benefit would have been acceleration and maneuverability (very important in combat) which, I guess is why it was so important to pack as many oarsmen as they could into the boat.

Edited April 29, 2022 by Richard Braithwaite

[Bedford]

I found those diagrams interesting but as a rower I have some different ideas about it.

 

When rowing with others, ie a boat with 4 oarsmen, you come to learn how the physics have to work.

 

If the oars are already on the power stroke before entering the water you're wasting stroke power and putting undue stress on your body when they take up in the water. Ideally you need to learn to engage the water just as the power stroke starts so it's a much more pronounced plunge into the water than a gradual arc. This way the effort ramps up quickly as the stroke starts, more oar in water = more effort, but doesn't come on after you start the stroke if that makes sense.

 

The return stroke should be pretty flat, why waste effort raising the oar through an arc? The power stroke should also be fairly flat once the blade is fully immersed, diving deeper wastes power. 

 

So I would suggest that the diagram of the movement of the oar handle is actually closer to the desired movement of the oar blade. Keep the directional arrows as they are but swap catch and finish, flatten the top of the stroke a bit and you've got a practical rowing stroke at the blade of the oar.

 

[Richard Braithwaite]

3 hours ago, Bedford said:

I found those diagrams interesting but as a rower I have some different ideas about it.

 

When rowing with others, ie a boat with 4 oarsmen, you come to learn how the physics have to work.

 

If the oars are already on the power stroke before entering the water you're wasting stroke power and putting undue stress on your body when they take up in the water. Ideally you need to learn to engage the water just as the power stroke starts so it's a much more pronounced plunge into the water than a gradual arc. This way the effort ramps up quickly as the stroke starts, more oar in water = more effort, but doesn't come on after you start the stroke if that makes sense.

 

The return stroke should be pretty flat, why waste effort raising the oar through an arc? The power stroke should also be fairly flat once the blade is fully immersed, diving deeper wastes power. 

 

So I would suggest that the diagram of the movement of the oar handle is actually closer to the desired movement of the oar blade. Keep the directional arrows as they are but swap catch and finish, flatten the top of the stroke a bit and you've got a practical rowing stroke at the blade of the oar.

 

I do take your point. Not the most efficient stroke...

 

Apparently the diagram was taken from a video record so should be reasonably accurate representation of what this particular oarsman achieved at this time. However, "the video record did reveal the variable quality of the bladework". Looking at the report which shows similar traces of what some of the other rowers were doing it appears that the diagram is one of the better ones (by a long way!). Bear in mind that this was 170 people rowing together for the first time in a vessel that they were unfamiliar with often from constrained seating positions with very little visibility. So technique was not optimal at all...

The report does comment on the large difference observed between the effective stroke length and the total length of the stroke saying that "the reasons why they mostly took their blades out of the water long before they had finished moving them sternwards require investigation.."

One reason suggested was that "the high moment of inertia of the oars which meant that they could not be manipulated quickly: if they were slowed prematurely because of this they would have to be taken out of the water early to avoid backwatering"

 

Even in a modern high performance racing 8 the blade is moving before it enters the water in order that it is at least travelling at the same speed as the water that is passing the boat (otherwise there will be a degree of backwatering at the catch and a negative force on the boat). For similar reasons the oar will be moving at some speed at the finish. 

The distance taken to accelerate the oar to this speed before the catch ("catch slip") or decelerate it after the finish (the  "release slip")   will largely be a function of the inertia of the oar (as well and the strength/skill of the oarsman) as suggested in the report, and the oars fitted to Olympias were much heavier and had higher inertia than those of a modern racing shell.

 

They did make some effort to address this with lighter oars in later trials on Olympias. This did enable higher speeds to be achieved, but I haven't seen any traces of oar path to see if the catch and release slip had been significantly reduced.

 

There are some good diagrams at the following link which shows this effect and oar traces for modern racing shells.

 

http://biorow.com/index.php?route=information/news/news&news_id=30

 

even these guys seem to waste some energy moving the oar up and down in the water during the power stroke

[Ian_Grant]

I built a jig to measure the force required from a sweep servo. It consists of 20 oars in a single tier as shown:

The clamps at the ends allow adjustment of the "gunwale" position to change the ratio of the oar inboard loom length to total length i.e. the "gearing" of the oars. I just cut a flat on the ends of the oar shafts. Rubber bands hold the blades in place, allowing me to try different blade sizes. I slapped some "Helmsman" spar varnish on them. The plan now is to borrow/steal a spring fish scale and use it to pull the oar looms to measure the force.

 

Still the cheap screw connections. For the actual ship I will use u-channel from Servocity to hold the oar ends. This channel is drilled at 8mm intervals. I will space oars every 5th hole, so 32mm apart. The upper tier will use the "3rd holes" in their beam to place them centred between oars in the lower tier. I was thinking of 3D printing a better oar attachment to allow them to pivot nicely in the u-channel. Something like this which I drew in TinkerCAD:

 

The small hole is for an M4 machine screw which passes through the sides of the u-channel. The large hole is for the oar shaft to glue in. Disadvantage is that once glued in I cannot remove an oar from the boat, through the thole hole. I even contemplate making brass oars, then I can solder the blades in and solder a square tube at the inboard end; in this case the larger hole in the fitting would be square, obviously, and I can perhaps drill another hole to pin the shafts into the fitting and allow oar removal.

 

The u-channels themselves will simply be hinged to the beam to allow them to tilt as the beam moves up and down. I have decided to use Servocity T-beam extrusions and mini v-rollers to mechanize the vertical movement, instead of round shafts and linear bearings. A bit cheaper, plus I can space the v-rollers at whatever distance whereas there are only two choices for length in linear bearings.

 

That's all for now. Eager to see what kind of force is needed!

Edited November 7, 2022 by Ian_Grant

[mtaylor]

If you want to remove the oars, think about a hole drilled through the top (the round area) where the oar fits.  Then do a lock screw through the hole and into the oar.  There's probably a better way but this is just off the top of my head as a simple way to do it.

[rookie]

This is heavy-duty engineering talk!

my head is spinning 

[Ian_Grant]

On 5/6/2022 at 4:45 PM, mtaylor said:

If you want to remove the oars, think about a hole drilled through the top (the round area) where the oar fits.  Then do a lock screw through the hole and into the oar.  There's probably a better way but this is just off the top of my head as a simple way to do it.

That would be ideal. My brother has a PLA printer; not sure I could drill and tap PLA without it cracking.

 

Maybe I could prototype it in PLA, finalize it, then get my local machine shop to CNC eighty out of aluminum for me, including a tapped hole for a grub screw. Now that would be awesome!

 

In fact, I can cut aluminum with my tablesaw/router using carbide teeth/bits. I'm sure I could jig up something to drill the holes repeatably....Thanks mtaylor!

Edited May 8, 2022 by Ian_Grant
garbled text

[Ian_Grant]

Finally, after all these months, some actual numbers!

 

Borrowed a fish scale from a neighbour. Testing in my (as yet unopened) pool, it reads about one lb pull when moving the twenty looms in a two inch stroke over one second. That's with blades measuring 5/8" x 1-3/4", and a loom length of two inches which is my desired value given the boat's likely beam.

 

I can extrapolate from there. Force will vary in proportion with blade area, and as the square of the stroke velocity (as I can attest from many many hours pedaling a bike into the wind!). Stroke isn't going to be faster than one second, and will be slower most of the time meaning much less torque (inverse of the square).

 

Each sweep servo will power two remes of oars, so 2 lbs thrust needed. If the servo only moves 60 degrees it will need a 2" arm to give the 2" stroke which represents a torque of 4 lb-in.  If it can move 120 degrees then a 1.15" arm will give the 2" stroke, representing a torque of 2.3 lb-in. Obviously a higher-rotation servo is desirable. Since fastest realistic stroke is about 1 second, servo speed is not likely to be a concern.

 

In an earlier video I showed return stroke speed at 1.5 times the power stroke speed which looked pretty good. If I stick to this value then the oars will be in a power stroke 60% of the time, and a return stroke 40% of the time. Sweep torque will be essentially zero during the return stroke. Sweep torque during the power stroke will vary depending on the angle of the servo arm , but let's just assume it is a constant 4 lb-in (to take the smaller-rotation servo case above) to be conservative. Over a cycle, then, the RMS torque will be about 3.1 lb-in (if I calculated that correctly). This value is the equivalent continuous torque. Since I can't get any info on continuous torque from Hitec, I'll need to do some temperature testing at this torque, with the servo arm going back and forth with an appropriate weight on it.

 

If I find a servo that does 120 degrees, then the 2.3 lb-in power stroke torque yields an RMS torque of just 1.8 lb-in.

 

Now I can look through servo specs and make an intelligent choice. I have been unable to pry any info from Hitec regarding continuous torque specs or thermal limitations so I will just have to select a servo whose max torque, "heavily" de-rated, is likely to indicate a continuous torque rating exceeding the above RMS values.

 

So, rookie, your head must really be spinning now!  😃

 

It's OK everyone, I can tease rookie......he's my brother!

 

Other unknown is what force is required to propel the actual boat? I just happen to have a WW I battle-cruiser I built for RC in high school which has a length, beam, and draft very similar to the proposed galley. I used the fish scale to pull it on the water. From a standing start it needs less than one lb to get going, which is much less than its displacement. Since four remes of oars should provide in the region of 4 lbs thrust, I think maybe I'm making a dragster! 😀

 

LATER EDIT:  I forgot to take into account the oar gearing. For the twenty oars collectively, the 1 lb pull from the servo at the looms yields 0.35 lb at the centroids of the blades. Four remes thus provide 1.4 lbs thrust against the water, far short of the 4 lbs I mentioned above.

 

Here are some pics of the battle-cruiser beside the oars at the proposed scale (fish scale is also seen). Size looks about right! The ship is Admiral Beatty's flagship HMS Lion, which was heavily damaged at Jutland. I got the plans from David MacGregor in the UK. Not a completely detailed build but pretty good for a teenager in the 70's. Powered by twin Decaperm motors (lovely motors, one of which remains) and oversized propellers which pushed it along like a destroyer. The other motor must be in one of my other old RC boats. There's even an old servo in there; might be fun to try it out.

Edited May 8, 2022 by Ian_Grant

[rookie]

very cool

great to see the old battleship again !!

does it still float ?   😬

[Ian_Grant]

Yes she still floats. I cleaned off years of accumulated dust for the picture. I got to thinking it might be fun to run her again, same decaperm motors but with modern battery and brushed ESC's instead of the old rheostatic speed controllers I used to use in her. Well do I remember cleaning the contact area so the copper wiper arm could touch the coils of the partially-uncovered wire-wound resistor element in the rheostats....  no one under 50 knows what I'm talking about...  Haha  😃.

 

I think I will consider refurbishing her and taking her to the cottage for the calm evenings. Providing that the other decaperm is in fact in one of my other boats.

 

I even thought it might be nice to pull the drawings out of the tube in the basement and add some more details. Her decks look pretty bare compared to some I've seen on MSW with hatches and scuttles all over. Probably I had trouble reading the drawings back then.

 

LATER EDIT: I pulled out the drawings and wow, did I ever skip a lot of detail. Her decks are indeed full of hatches, ventilators, and scuttles. Also omitted all the light armament, armour belt edges on hull, protrusions on turrets which I assume are rangefinding lenses (?), vent pipes at the stacks, the frames at the top of the stacks (for covers when laid up in port?), all the doors, many many boats, etc etc. ill consider fixing her up but life is short; I still want to do this galley first.

Edited May 22, 2022 by Ian_Grant

[Ian_Grant]

Servo selection done - the sweep servos will be Hitec's model HS-755MG, a "giant scale" or "1/4 scale" (depending on who you talk to) analog servo where "MG" stands for metal gears. The other contender was the HS-645MG standard-size servo, which would have had less margin over the expected RMS torque. The 755 rotates 90 degrees with the typical RC set's 1000 to 2000 usec pulse range but can rotate up to 200 degrees given a wider range (570 to 2400 usec is the extreme, according to Servocity). Since I have the Arduino controlling it I can easily provide 870 to 2130 usec or whatever, to get it to rotate 120 degrees. Then as discussed previously the arm can be short, minimizing the torque.

 

The final nail in the 645's coffin was that the 755's different motor actually draws less current while providing 50% more torque. I think I said earlier that I'd rather have a bigger motor loafing along than a smaller motor labouring; with the bigger motor drawing less power it's a no-brainer.

 

I then needed to design the oar position angles for power and return strokes. Pulling out my Pitassi I was reminded that his quadrireme interpretation had 88 oars, not 80, so my remes will be 22 not 20. This will increase torque 10% but I had already decided on smaller blades for appearance so torque is only about 3/4 of what I said in earlier posts.

 

I spent some time manually drawing the oar angles wrt the waterline given the outrigger size. I found that it's not easy to have both remes of oars in the water for the power stroke and still have decent clearance above the water for the upper reme of oars during the return stroke; in a real ship the rowers in the upper reme could just lift their blades more than their buddies in the lower reme, but my mechanism will move the looms of both sets of oars the same distance vertically. It seems I will need to have the lower reme come up to horizontal in order to have clearance for the upper reme. Also, I'd like to have a little more freeboard than scale (which is only 3/4") but this makes the oar angles steeper which increases the above difficulty. I finally came up with a drawing with lower oar hole centres 1.25" above the water, with upper oar tholes spaced 13/16" higher, and 1.25" more outboard. This will lead to an upper deck somewhat higher than ancient writers imply for a quadrireme, but what the hell. I was surprised to find that I only need about 1/2" of vertical movement at the looms. Here is the drawing:

 

I could go on forever making manual drawings so I decided to use tinkerCAD to model the mechanism with Servocity parts. I still can't figure out how to import Servocity's STEP files into tinkerCAD, and I don't want to spend months swearing at Fusion360, so I just drew the parts manually in tinkerCAD. Still required a few special words. 😁

 

I drew it with linear bearings since the available 2" long bearing now seems about right. My main purpose is twofold: (1)  see how far below the waterline the drawer slides need to be, in order to design the central bulkheads, and (2) determine how far towards hull centreline the mechanism reaches i.e. how much space in the middle (needed!) for battery or servos?

 

Here's what I have to this point. This drawing shows two upper reme oars and one lower reme oar with blades in water. Also two lower reme oars in horizontal position. Oar angles when in the water are 17 and 26 degrees respectively, a bit higher than Pitassi's 12 and 19, because I have the extra freeboard.

From the bottom, we have the red drawer slides, a yellow "low side channel", an orange clamping hub, a blue SS shaft clamped in it, a green linear bearing, white clamping mounts around the bearing, a blue "back plate" which forms one end of the oar beam and which slides up and down the shaft thanks to the bearing, purple hinges connecting yellow u-channels to the beam, orange oar pivots within the u-channel, and orange oars. I haven't represented the other end (similar), or the L-channel beams which will connect them. The hinges on the upper reme will connect to the beam via standoffs (not shown) which will place the upper oar looms correctly relative to the lower (given the outrigger dimensions).

 

So far I've concluded that the bottom faces of the drawer slides only need to be about 2mm below waterline, not as much as I had feared. The mechanism intrudes about 2.6" into the hull meaning that I only have about 1.8" free in the middle of the 7" beam hull. Probably ok for a battery and maybe the lift servos.

 

I'm also finding that it is not so easy to place the second reme's hinges exactly as required; I had blithely assumed that if I tweaked the spacing from beam to hinges, and used washers as necessary between hinge and u-channel, I could get the oar looms to be anywhere I liked. Not quite that simple.

 

Further tinkering required. It would be cool if I could animate this. Dream on.

Edited May 30, 2022 by Ian_Grant

[Ian_Grant]

Had a bit of a brain wave: how about two independent beams for the upper and lower remes? This would allow software to independently control the range of lift motion for each reme of oars, allowing the upper ones to be raised slightly higher without the lower ones necessarily going all the way up to horizontal. I like it!

 

Ideally the vertical shafts could have two shorter linear bearings on them, one for each reme's oar beam. I would modify the interface board and software to add two more output servos for the two new lifts. Potential snag is we now have a long beam mounted on two 1.25" linear bearings more than 2 ft apart; could a single central servo move it up and down without the ends binding? Don't know. I have always had in the back of my mind that perhaps the beams would need servos at each end, so be it. But with four beams that would be eight servos doing the lifting which seems beyond the realm of reasonableness.

 

Or, perhaps, two vertical shafts at each end, the inner pair for the lower oar beam, outer pair for the upper, and all four shafts sporting 2.25" bearings to better resist binding.

 

I am going to order some parts from Servocity to build a partial model, and get a "feel" for how these bearings perform in my scenario. If central lift servos can't work, then back to Plan "A".

 

More to come.

 

Happy modelling.

Edited May 30, 2022 by Ian_Grant
clarity

[starlight]

Hello Ian,

 

This is a very interesting subject and I must say you've done some very impressive work so far!

 

On 5/21/2022 at 11:08 PM, Ian_Grant said:

Further tinkering required. It would be cool if I could animate this. Dream on.

 

I am fortunate to have a student license of Solidworks, which has an excellent motion study and rendering suite. I think recently Dassault released a "maker" license available at low cost. There is also Fusion 360 which I think has the same capabilities but I don't have any experience with it. Certainly, some sort of CAD program where you can verify ranges of movement and linkage designs might be very useful.

 

Cheers,

starlight

[Kevin-the-lubber]

Ian, I don’t know how far you’ve gone with F360 but it’s tailor-made for this kind of thing. I animated the moving parts on that little serving machine I made the other week, just to check that I hadn’t made any stupid design errors. But it’s also a lot easier than Tinkercad for creating and editing the parts you have there. I cut my 3D teeth on Tinkercad and still think it’s quite under-rated but, to be honest, can’t see myself ever using it again.

[Ian_Grant]

Oh Kevin, I haven't tried 360 at all.......too much of a learning curve, and I get frustrated and angry at computer tools with their idiosyncratic user interfaces. I have better instincts for physical tools and power equipment.

 

At this point as I said I am going to order some parts and do a partial prototype build to check how the vertical movement would function, and just generally what these components are like.

 

LATER EDIT: Parts order placed. My heart nearly stopped when I saw the $US total, never mind in $CDN. I took the two sweep servos back off the order to make it more palatable; after all, who knows, in the end I may decide to give up!

 

As an aside, one of my neighbour's large maple trees was felled in a recent bad storm. Unfortunately it landed on and crushed our above ground pool! Preliminary talks with our insurance informed us that they will only give us a pro-rated value for our admittedly old pool, perhaps as low as 25% 😭. Yes the pool was old, but it was fine and enjoyable and I wish they would just make it good for us. We could be on the hook for several thousand dollars to replace it at the moment,  meaning the Admiral might find it  of debatable worth. It's mainly me that enjoys it 😳 now that the kids are grown up. That's why I'm a bit sensitive to expenditure at the moment.........

Edited May 30, 2022 by Ian_Grant

[Ian_Grant]

17 hours ago, starlight said:

Hello Ian,

 

This is a very interesting subject and I must say you've done some very impressive work so far!

 

 

I am fortunate to have a student license of Solidworks, which has an excellent motion study and rendering suite. I think recently Dassault released a "maker" license available at low cost. There is also Fusion 360 which I think has the same capabilities but I don't have any experience with it. Certainly, some sort of CAD program where you can verify ranges of movement and linkage designs might be very useful.

 

Cheers,

starlight

Starlight, thank you for the encouragement. When I decided I wanted to do an RC build again I was torn between the ideas of a functioning galley and a square rigged vessel. I thought a galley would be more of an engineering challenge, but boring to use once the novelty wore off; whereas a ship would require less challenging preliminary work but be a constant challenge to sail. I opted for the galley now, and hope to do a square rig later!

 

It's been fun playing with electronics again, and I actually quite enjoyed writing the software. I know now that my code is pretty basic and I'm looking forward to improving the stroke's realism. There's plenty of program memory left in the micro! 😃

 

Speaking of realism, could they even back-stroke in the real thing?? 🤔 It would be harder to do than when they tried to evade "Talos" in "Jason and the Argonauts" ....  love that movie! 😀

Edited May 30, 2022 by Ian_Grant

[Kevin-the-lubber]

If and when the mood grabs you, I’d really recommend this DIYODE article https://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwjQuu2YmYf4AhXNYMAKHb6eB_QQFnoECA0QAQ&url=https%3A%2F%2Fdiyodemag.com%2Feducation%2Fexploring_3d_part_1_beginners_guide_to_fusion_360&usg=AOvVaw1q5hCjynXrbTUKYaEWwVl7

 

Whichever way you go at it (CAD), nothing will be as intuitive as Tinkercad, but this tutorial was pretty quick, easy to understand, and saved me from throwing the mother of all computer generated tantrums.

[Ian_Grant]

4 hours ago, Kevin-the-lubber said:

Whichever way you go at it (CAD), nothing will be as intuitive as Tinkercad, but this tutorial was pretty quick, easy to understand, and saved me from throwing the mother of all computer generated tantrums.

"...saved me from throwing the mother of all computer generated tantrums"  HAHAHA! 😄

 

Thanks Kevin, I'll take a look at the video.

[Ian_Grant]

Just a note to myself so I don't lose the figures.....

 

Plotted some more parts in Tinker; in order to have two 30mm linear bearings on a single shaft for separate upper and lower oar beams, the drawer slide bottoms need to be 21mm below water level to prevent the two bearings interfering on the shaft. This allows upper oars to move over a greater range than the lower, while still allowing lower oars to go up to horizontal if required. Shaft length is 95mm.

[Ian_Grant]

WARNING:  Rant on US/Canada border trans-shipping ahead..........

 

Well I got my package from Servocity today. If I thought my heart nearly stopped when I saw my Cart total, it came even closer when the UPS guy wanted another $109CDN from me at the door for, ahem, GST and brokerage fees and taxes on brokerage fees. This on top of already paying $49.44USD for shipping $188.57USD worth of parts to the border.

 

Servocity has a standard shipping fee of $8.99USD within the USA, so a US resident of Ogdensburg, say, gets his/her shipment for $8.99 but I pay $40USD more for my shipment to travel the extra 6km to the border. Then add $69.30 for brokerage fees and $9.01 tax on that. Finally add Canadian GST which is also applied to the shipping fees not just the product value.

 

Bottom line, I received $188.57USD  ($236.83CDN) worth of parts for $408.28CDN to my door, an increase of merely 72.3% in shipping and up charges.

RIDICULOUS! HIGHWAY ROBBERY!

 

And that just allows me to PARTIALLY build ONE SIDE of the galley mechanism.....at these prices I may use these parts as drill guides and revert to a mix of metal and wood components, or maybe just give up altogether.

[rookie]

This is Canada

Everything is more expensive.... 

That is how Canadians like it, apparently   🙄

[Kevin-the-lubber]

Wow. That's what the UK used to be like way, way back. I once had to pay about £100 customs duty on a £200 guitar I bought on a trip to California; I could have wept, especially as I couldn't even really play the thing. 

[James H]

18 hours ago, Ian_Grant said:

Servocity has a standard shipping fee of $8.99USD within the USA, so a US resident of Ogdensburg, say, gets his/her shipment for $8.99 but I pay $40USD more for my shipment to travel the extra 6km to the border. 

 

Do you know anyone just inside the US border who would receive for you, and you could pic up from them? Is that possible between the US and Canada?

 

 

26 minutes ago, Kevin-the-lubber said:

I could have wept, especially as I couldn't even really play the thing. 

 

I know that feeling! 😆

[Paul Le Wol]

Hi Ian, I realize that this doesn’t help your present situation but, whenever I order something from offshore, I try to find a retailer that offers a postal service option for shipping. Say for example USPS. Even if it initially costs more, you won’t be subject to brokerage fees. Occasionally your shipment may get flagged and you will have to pay GST.  I’ve only had that happen a few times. At least this has been my experience.

Edited June 9, 2022 by Paul Le Wol

[Ian_Grant]

So I assembled a bunch of parts into the required length of mechanism.

The beam is a bit of a kludge of disparate parts, and the baseplate is longer than necessary because the shorter length was out of stock. Shown are the 2-1/2" linear bearings which do slide nicely on the SS shafts. You can see how lo-o-o-ng the beam is compared to the length of the bearing surfaces at the ends. In point of fact, it cannot reliably be operated by a single servo in the middle (or in this case my finger); one can see sometimes different vertical shifts at the two ends. So the scheme with two beams on only 1-1/4" bearings is out. Probably just as well. There will be a lift servo at each end of each beam, with the servo pair on a given beam y-harnessed to the Arduino output.

 

The U-channel for mounting oars has to be in three pieces as shown because this channel only comes in limited lengths.

 

I also find that the hinges are stiff, too stiff in my opinion. Oil seems to make no difference. Can I find some piano hinges with metric hole spacing I wonder?

 

Still wondering how much of this stuff to use, and how much to make from Baltic Birch plywood.

[rookie]

that is some high-end Meccano.... 

[Ian_Grant]

Yes it could reasonably be used in parts of a planetary explorer ship 😄

[Glen McGuire]

Just found this build log, Ian.  Incredible work so far on a unique and fascinating project!  Look forward to following the rest of the way on this.