Test results from prototypes

I've made a shaft to Design 2, solid hardwood (ash) inboard and solid trapezoidal outboard from Western Red Cedar, 4.3m long. I've left the outboard a little oversize for trimming and tested for UHF and stiffness. UHF with the blade where it should be is 3.2kg, still a bit high. Deflection at 2.5m with 10kg, supported at the middle gear of 2.8 true gearing is 60mm. Since this is stiffer than our target of 80mm I'm going to shave bit more off the outboard. Total weight about 5kg at the moment without varnish or the keeper.

I'm pleased it is so stiff. The higher than expected UHF is partly due to the cedar being denser than I bargained for. Published figures which I used in the spreadsheet are 370kg/m^3 for which UHF was predicted to be 2.2kg but samples came out at 420kg/m^3 which in the spreadsheet gives 2.5kg.

I'm away for a few days but when I return I'll take the shaft down to the designed dimensions and test it again. I'm quite encouraged because it's the best combination of UHF and stiffness of any solid oar in the master spreadsheet, and about the same as Don's hollow radiata oars.

Further to the last post, the shaft is trimmed down and the deflection has only increased to 62mm. Now that I've cut out the curve at the end of the shaft to take the blade, the UHF with blade is down to 2.6kg, pretty near the design UHF at this density of cedar of 2.5kg.

Total weight is nearer 4kg than 5kg, but I've become aware that my cheap spring balance is not all that accurate. The UHF and deflection were measured with more accurate means so I'm happy with them.

I need to sand and varnish this oar, and add the wear strips and keeper, so it will be a week before I can row with it. But just pushing down on the handle it feels nicely light.

The 4300 long ply web test oar is complete. Missed all the numbers by miles, but put that down to my conservative oar building nature. I set out to get it as close to the true gearing of our existing oars as I could, because that would give a better comparison - ie rowers would not be feeling changes that were simply the result of a different gearing. True true gearing (!) was measured at 170mm in from handle end (normal mid point of hands), and centre of area of blade. This produced a FISA gearing of 2.825, so it's within the normal range. All other figures are as per the measurement form. Numbers as follows based on FISA gearing of 2.825
Length 4300
Inboard 1124
Outboard 3176
Oar width at pin 74
Oar depth at pin 56
Width at neck 42
Depth at neck 30
Blade length 600
Max blade width 220
Total weight of oar (plate and spacers not fitted) 4.6kg
UHF 200 in 2.47kg
point of balance from pin 492
deflection at 2.5M from pin with 10kg, 49mm.
Blade is 4mm ply (same as webs), grain vertical, area is 1150 sq cm, exactly the same as our previous blades, once again to keep comparisons valid. Flanges and filler/handle are Alaskan Yellow Cedar (just 'cos I had a stick of it), density of actual stick was measured prior to starting, 504kg/cu M. So only a gnat's whisker above the mean for the species, so MOE should be around 9.79 Gpa.

Filler/grip is 1410 long, 100mm long birds mouth at inner end.

So it's a stiff little critter, about 12% lighter than a comparable oar with 6mm pine top and bottom. UHF is down about 30% from my previous best. I used a 15mm flange on the compression side, tapered to 8.5mm at the neck. Flange is 11mm on the tension side, tapering to 6.5mm at the neck. Vertical stiffness seems OK, so it looks like the ply web idea has legs, and a braver person than I should be able to eek out a bit more from the weight and UHF figures, albeit at the expense of a bit more bend.

It should be a cheap enough build - a set of 4 would come out of 1 sheet of 4mm ply (2500 by 1220), 1X4M length of good quality 150X50 timber for the webs, and one 3M length of 100X50 for the filler/grips. There's obviously a bit of wriggle room for the dimensions if you were happy to accept a bit more bend.

It's going to be pretty tough to get a UHF down to 2kg, and you're going to end up in a tough old trade off whereby if you try to reduce weight inboard the UHF will start going up again. As noted there's a bit of room for improvement in the UHF - if you went to a 1000sq cm blade you should chop 100 grams out of the UHF, and you might wring another 100 grams out of the flanges, but they are going to end up frighteningly thin! But on the face of it ply webs look like a practicable proposition, and even as built it feels significantly lighter than our normal oars, and better balanced. I'll splash some varnish on it, add the rowlock plates and report back on its performance in use.

Great work Don! It is very stiff, which is a good sign. The density of the Alaskan Yellow Cedar is quite a bit higher than Western Red Cedar, so that's going to make the UHF more, and it's less than oak or ash, so the inboard will be lighter than a hardwood one.

I see why you made the compression flange thicker than the tension one but as far as I can see we are designing for stiffness, not strength, and if the compression flange was 11mm like the tension one it would save a bit of weight. I'm going to do that and WRC is weaker than AYC so it may break!

I feel that the 2kg UHF is a sort of Holy Grail but I think we'll get there. My solid oar here has ash@770kg/m^3 for the whole inboard and WRC@420kg/m^3 for the outboard, deflection 62mm, UHF 2.6 or 2.7. I think from hefting it in the workshop it is as soft vertically as one can get away with, so it may represent the peak of what one can do with solid oars.

I'm going to start one like yours but with ash, cedar and 4mm okoume ply. I'm also going to end the flanges and the webs just outside the pin, with scarphs on to the ash inboard, so the entire 75x54 cross section is ash. At 420kg/m^3 for the cedar the UHF should be 2.0kg.

Anyway I'm keen to hear how your one feels to row with, and next week should be able to try my solid one out.

Do you think there is an advantage in bigger blades than 1000cm^2? Trading off windage against slip.

Oh and Rory Cowan on Arran is going to make a hollow outboard out of all WRC (no ply) so we will have three hollow and one solid oars to compare. Then we'll have to decide which is worth making a set of 4 out of.

Lots of stuff there Topher. Re the flange thickness, there were a couple of reasons why I went for the 15mm compression flange. 1) I'd assumed that compression vs tension performance was roughly linear up to failure. I don't know for sure, but 2) using the thicker flange helped minimize waste for the timber I was using. I've been happy to wriggle component dimensions around a bit to suit standard timber dimension. That's why the width ended up at 74mm, not 75. And blade length 600.
That way I could cut 450mm off the end of the sheet, and get 4x600 long by 220 wide blades (grain vertical), then get 16 strips of ply out of what was left. Scarfed together that got me enough for the webs.
I took the depth to 56mm simply because the flanges were ripped out of 50mm rough sawn timber, so ended up 48 deep, and given our concern about whether the oar would be stiff enough vertically, that's what I left 'em at!
Given the objective of coming up with a cost effective oar, I was happy to go with 74mm width to avoid having to do fancy tapered cutting of the webs.

Re blade size - that's just one of life's mysteries! After talking to Finlay a few months back I started to experiment with larger blades and lower gearing. The feathering square currently sports a 1350 sq cm, 550 long, and 355 wide "pizza oven spade". It's butt ugly, but generally nice to row with. With a feathering oar you can just stop worrying about all this windage and blade shape malarky and just make what works for the pull.
That said, 1350 sq cm and a stiff loom (49mm deflection@2.5M) is definitely too much for cruising. If the boat is stopped by a wave, the oar is tough on the shoulders whilst you build speed again. But in the hands of a young and fit person it would be a real weapon in a race.

We started out with 1290 sq cm blades, thought they were a bit big, so cut them down to 1150. That seemed like a nice size so we stuck with them, and for the new oar I just shortened the blade to 600 then fattened it up until I was back at 1150 sq cm. It's not the most elegant shape I've ever seen, but it's not so ugly as to send small children screaming from the shed!

I just bite my lip and move on from all the posts about blade shape for clearing waves, windage etc. It just isn't part of my consideration. I made plenty of non feathering oars for our Raid last year, but to my mind non feathering oars are just a cheap and cheerful stop gap for getting people into the boats and enjoying being out on the water. But rowing around in 20 knots of wind with a non feathering oar is like trying to pull screws out of a piece of timber because no one told you that you could turn them to get them out!!

I'll send photos in the next day or so.

Thanks for that Don. I see the way it all fits in a sheet of ply, which is neat.
It grieves me not to use your elegant feathering system but currently I think it may be too much complication for the builders and too much of a culture shock for the rowers. This may be a lack of moral fibre on my part.

I like the pulling out the screw analogy!

There is an old Scots proverb "a fool and a bairn should never see a job half done" but since you guys are neither fools nor bairns here are come photos of the solid oar.

I glued the blade to it:


but didn't like the wedge which I had to insert at the root end of the blade, thought I could save some weight, did it again by bending the forward strip:

This did in fact save weight, and without varnish the UHF is: at 3.0 true gearing, 2.86kg, at 2.8, 2.43kg, and at 2.6, 2.03kg. Because the CofG is now so close to the pin, the small movements between gears are making a big difference to the UHF. I think we should use the middle value, at gearing 2.8, if we are comparing oars.

That's a nice looking blade Topher. I think you'll have to put some thought into how you control the shape from the one design point of view, but it certainly passes the aesthetics test. I know we went round and round a bit deciding on how we'd measure true gearing, so just 'cos I'm feeling lazy could you please confirm the system you're using for measuring true gearing, then I'll re measure the ply web oar just to ensure comparisons are valid.

The slightly arbitrary standard is 150mm in from the end of the handle to the centre of the pin = inboard, and centre of the pin to centre of area of the blade = outboard. True gearing = outboard/inboard.

For long sea blades we just take halfway up the blade as the centre, for macons tulips and needles it's a bit further out. I could do a more accurate calculation for the current blade.

I've been making a three gear set of "Dawson Dimples" and will post a picture in the oarlock section. It may not work but it sure is fascinating to make.

For a 2.8:1 true gearing I now get a 1167 inner, and UHF of 2.265kg. Didn't do a UHF for 3:1 because that will be close enough to 2.5kg to not be worth worrying about, based on the previous measurement. I think if I use 13mm dia spacers I'll be able to do a vertical plate and spacer rowlock with 2.6:1, 2.8:1, and 3:1 gearing. The interval between 3:1 and 2.8:1 is only 51mm, but with narrower spacers it should give adequate vertical movement. Getting 3 ratios for the stroke oar could start getting tricky, especially using a 28mm thole pin. Anyway, that's for later.

Hi Don,

We are now getting close to the targets. 1167 inner, with say 70 more mm from an inrigger, should be plenty to avoid clashes even at 3.0. I see what you're saying about the gap between the gears getting small, especially at stroke. The mechanical demands on the spacer are becoming almost too much for wood, but we can think about that later. There is quite a lot of resin in mine, see the mechanics of oarlocks thread.

Today was to be the first row for the Design 2 solid oar but frustratingly it was too windy. When I sit in the boat on the trolley pretending to row it feels nice, and when I walk down the pier and pretend to row in the sea, it floats at the right depth and grabs the water strongly.

Pretending to row with my new oar, it's a bit pathetic, no? I should get out a bit more.

I thought the Scots came up with the saying "no such thing as bad weather, only bad clothes".  Perhaps your boat (and/or oar) is not properly dressed?  Anyrate, the new ply web oar is ready to go down here, weather forecast is good for tomorrow, so with a bit of luck it will get some use tomorrow, and I'll report.  Couldn't get photos to post tonight, so have sent them via email to Topher.  Managed to get 3:1, 2.8:1 and 2.6:1 gearing on the new oar by using  "wasp waisted" spacers, and a 25mm pin.   Weight is up to 5 kg with rowlock and wear strip top and bottom (makes it ambioarstrous) to testing purposes.

We await the rower reports with interest! Can you fill in one of Rory's feedback sheets please?

Finally got out today, in pretty nice conditions. The oar was really light to row with, compared with our existing ones. The catch was positive and when I tried to take a stroke with the blade only half submerged, the vee of the bottom half pulled it right in, so it automatically goes to the right depth.

There seems to be more spray when it comes out of the water, perhaps entrained by the flat top of the shaft where it meets the blade.

In a 15mph headwind there felt to be a slight vertical vibration when the blade was out of the water, so perhaps it is shedding vortices in the air? Certainly there is no sign of vibration in the water.

The ball ended pin and keeper worked well, with no play at all and smooth in action. I was able to change gear while rowing, although you need to be quick.

The oar felt stiff in bending with no sign of vertical instability, so the shaft design is stiff enough.

Three other rowers had a try, all women, and they liked the light balance force. They also liked the blade, saying that it allowed a powerful stroke.

All in all, a very encouraging trial. I will continue to row with it in different conditions and report. Now I need to start on the hollow oars.

Prototype Oar and Rowlock

I understand that so far there is one solid and one hollow prototype oar made and a further two hollow oars under construction all of which are 4300mm long.
I am interested in trying to make a longer hollow oar with a slotted hole and rowlock as per the sketch submitted in the Mechanics of Oarlocks section on 23rd Dec. I have a few queries before I start :
1. The photos of Don's oar appear to show the flanges fitted
between the plywood webs whereas the design calculations seem
to have the plywood webs fitted between the flanges. The way
Don has done it would be easier to construct and I wonder if the
calculations have been carried out for this version.

2. Are oars longer than 4300mm still being considered as a standard
design option.

3. In view of the comments made elsewhere regarding rowlock pins
made of stainless steel is there any point in trying a prototype as it
would appear they are already ruled out. What is the consensus of
opinion?

Hi Robert,

Great that you are interested in making prototypes. Lots of questions, so in the order you ask them,
I am interested in trying to make a longer hollow oar with a slotted hole and rowlock as per the sketch submitted in the Mechanics of Oarlocks section on 23rd Dec.
The sketch shown in my post of 23rd December was of a full oar section on the left with a separate keeper on the right of the pin. The slot was formed between the keeper and the forward face of the oar. This is the system I used on the solid oar just completed, and so far has performed well with no play at all. However I plan to make the next oar with a slot cut in the oar inboard itself, capped by a plywood plate with three dimples.
1. The photos of Don's oar appear to show the flanges fitted
between the plywood webs whereas the design calculations seem
to have the plywood webs fitted between the flanges. The way
Don has done it would be easier to construct and I wonder if the
calculations have been carried out for this version.
I'm embarrassed to say that you have found two flaws in the calculations. You are correct to say the calculations work as though the ply webs fit between the flanges whereas it should be the other way around. I have corrected the errors and I have increased the thickness of the flanges to 16mm at the gunwale to make up for the lost depth, and as a result the UHF has gone up from 1.8kg to 1.9kg. Agreed this is much the easiest way to make the oar.
2. Are oars longer than 4300mm still being considered as a standard
design option.
Don and I made the judgement that a shorter oar was worth getting, even at the cost of making a block inside the gunwale for the pin at positions 2 and 3. This is a trade-off and definitely debatable. If you think a longer oar is worth making then go for it, and the SCRA will pay for the materials. If you want to make a 4.5m oar then Design 3 would suit (UHF 2.1kg), or you can make your own design.
3. In view of the comments made elsewhere regarding rowlock pins
made of stainless steel is there any point in trying a prototype as it
would appear they are already ruled out.  
You and Don both advocated this idea and I think it should be tried. I am advocating plastic pins with tapers going into tapered holes. Since this is my invention I don't think I should make the judgement about which is best.
What is the consensus of
opinion?
This is a really good question and not easy to answer. We need to decide between two hollow and one solid 4.3m oars, and your longer one, plus we need to decide between Don's pin and keeper and my ball ended pin, and between your stainless tube on a plate and my plastic tapered pin. Ideally we would all  gather round all these oars and oarlocks, try them out, and discuss them. As we are scattered all over the globe the best we can do is photograph, measure and film them, and then ask the whole team to judge. What do others think?

Robert, I can confirm that the NZ oar has the ply webs glued over the top of the fore/aft flanges. Because I was a lazy little beggar at school, I have to work with the TLAR design algorithm when developing oars (That Looks About Right). So I just fattened up the flanges a bit and glued the webs over the top 'cos that was the easy way to build 'em! We tried the new oar out yesterday, and the comments were very favorable - it's light to lift, UHF is noticeably less, and when you pick up the cadence downwind that lightness shows up in a nice quick response. Earlier on in the discussions there was some talk about measuring polar moment of inertia, and as I make more oars I can't help feel that just sucking weight out of (especially the outer end) is a key issue. Outboard weight drives polar moment of inertia, and the longer the oar is the more every gram counts, and the heavier the oar needs to be. And the higher the polar moment of inertia the less well an oar responds to higher cadences.
A year ago or so one of the St Ayles rowers here introduced me to a boatbuilder who had been the premier oar maker in NZ for the sliding seat brigade back in the days of wooden oars. My friend was enthusiastically explaining about the 4.5M oars we were making. The boatbuilder's face was a picture - he just couldn't hide the fact that he thought we were, top to bottom, left to right, barking mad! He just couldn't wrap his mind around why anyone of sound mind and body would build an oar that long. All he could see was weight and flex.
And over the last year I've become a believer too. I just can't see any virtue at all in length. As to the specifics of where you focus your developments, I can't comment with any authority, but from what I've seen over here, and had confirmed with the latest test oar, I think the most productive area to focus on is weight reduction. And shorter is lighter!

Hi Don
Wouldn't be the first time I was told I was mad!
I understand what you are saying about polar moment of inertia, our solid oars average 8+kg but that is compensated with a low UHF (average 2.5kg). There are a number of benefits with long oars in the St Ayles such as staggering the rowers and reducing the clashing problem, reducing the stroke arc to concentrate on the most efficient region perpendicular to the boat and reducing the vertical movement required to get the blade immersed.
I also believe that rowers can perform better over an extended period of time if they set a steady stroke rate that matches their breathing and make the boat go faster, not by increasing the stroke rate but rather stretch further forward and back more and increasing the power. This of course is greatly assisted with a low UHF.
The weight and UHF figures for the hollow oar look very interesting but I think they will be a challenge to achieve. It is only by pushing at the boundaries of the design will we find out what the limitations are, and then we can scale back if necessary.
That is why I wish to pursue a long oar.

Hi Topher
Had a look at the reconfigured stress calculations for the 4.5m oar with full width plywood webs and some figures are looking high. I think you mentioned somewhere about being able to play around with the figures but I cannot find that version. Is it still available for anyone to use.

Hi Robert,

You are right, the stress figures go a bit high towards the tip. You can alter any of the numbers, at least I can, and I think you could.

Playing with it now, I find that making it wider at the tip (increase cell P26) helps a lot. I've increased that to 35 and we now have lower peak stresses, have a look. But feel free to download it and play around with it.

UHF depends a lot on wood density. I assumed that Western Red Cedar had a density of 370kg/m^3 but the stuff I'm using at the moment is 420. It will be stronger and stiffer too, but I have not put that in the spreadsheet. What timber are you planning to use?

I got a piece of WRC at 349kg/m3, a plank of oak at 835kg/m3 and a sheet of 4mm marine ply at 587kg/m3. The wide variation in timber density makes it more difficult to legislate for a standard oar design.

I like your idea of starting the webs and flanges outside the rowlock pin as it means the oar design can be in two halves, outboard optimised for stiffness and lightness, inboard beefed up and weight distributed for balance. Have you a design detail of the flange and web joint area?
Tried playing with the spreadsheet but not getting the numbers I expect for weight and UHF. The programme is designed to carry the lighter weight flanges and webs inboard to the handle which is reducing the counterbalancing effect of the heavier oak. I am having a go at doing the inboard calculations manually.

Hi Robert,

I did at one point have that spreadsheet with solid oak inboard, and have again. I have put in the densities you quote, and the UHF is now a spectacular 1.5kg! You have found some very dense oak and some very light WRC. Because there are now two versions of Design 3, this is a link to the most recent one: https://www.dropbox.com/s/mbef8k9u87lu8s0/design%203%20%204.5m%20oar%20%20hollow%20outboard%20wrc%20flanges%204mm%20okoume%20ply%20webs%20%28topher%27s%20changes%202017-01-29%29.xlsx?dl=0

I'm having all manner of problems with posting drawings and photos on this sitte at the moment.  I sent the email below and a drawing to Topher by email yesterday.  Perhaps Topher could paste the drawing into this.  I'm just heading away for a few days so apologies for the hurried response.

Yeah, that's a pretty interesting idea.  I'd not thought of doing it that way, but on the face of it you could make some worthwile gains in reducing material waste.  Assuming you were using plywood webs there's no real gain in stopping them short - all you do is end up with more ply waste, more scarphs, and having to buy thicker timber for the oarlock section. So if I were doing it I'd go ahead with stopping the flanges short, as per the drawing attached, then run the ply webs over the top of the whole thing, all the way up to the base of the grip.  Benefits are quite interesting - assuming 150 by 50 timber is a common size (6 by 2 inches would be better) you'd get 2 grip/filler pieces out of each stick, and your flange material could come out of a single length of 150 by 50 of 3.1M.  In practical terms you'd probably have to move to about a 72mm oar width to allow for the kerf when ripping the 150 by 50 timber, but we've got a bit of wriggle room there as the prototype is plenty stiff.  You could glue up the filler/grip and flange pieces into one assembly, then glue the ply webs top and bottom later.  It would be an easy oar to make, with no big scary glue up operations.
I've done the drawing based on a filler/grip piece of 1410 length, simply because that's what I settled on for the NZ prototype.  At lowest gearing the pivot point comes scarily close (183mm) to the end of the filler, but so long as you glue a wear piece on to the compression side of the oar the structure looks OK.  Essentially the front wear strip becomes a structural back up, but that seems sensible to me.  Something like 700 long, with tapered ends should be a nice bit of insurance.  As you say, the scarphs will have to be pretty good, but I'd go for 12:1 (or longer) scarphs, and do the old Gougeon Bros trick of priming the glue faces with neat resin to get good penetration, then 20 mins later applying the glue mix then clamping the joint.  


I think your suggestion is a really nice step along the way to meeting the low cost part of the objectives - we are starting to get down to a build list that is all 150 by 50mm, one stick of 3M for the grip/filler pieces, one stick of 3.1M of fancy stuff for the flanges, and one sheet of 4mm ply for the webs and blades.
So yeah, perhaps run the numbers again for a 72mm width and see if they are OK, then do your prototype.

I'm about to glue up the first part of my hollow oar and am now pretty sure of timber dimensions. The ash inboard is 75x46 cross section, with the 4mm webs glued to the outside of that making the cross section at the pin positions 75x54. The webs stop at the inboard end of the pin slot and do not extend to the handle. This means they need to be 3250 long, so two strips 1650 long would allow enough for a scarph.

Since the webs taper, one strip can be 80mm wide including saw kerf, and one can be 70mm which means we can get 8 of each cut out of a sheet 1220 wide and still leave more than enough for the blades.

The inboard can come out of 50mm sawn timber, but 150 wide would not be enough to get two pieces 75 wide, so maybe we need 50x100 or 50x175 timber. Each inboard needs 1350 length.

The cedar for each oar is two strips of max thickness 12mm tapering to 3mm, 46 wide tapering to 17 wide. The most commonly available size is sawn plank nominally 25 (nearer 22) by 150. It is possible to rip this into three strips finishing at 46 wide, so we can get at least 3 strips per plank.

It seems wasteful to thickness 22mm down to 12, and in theory it ought to be possible to thickness down to say 19 or 20 by 46 and then saw the piece into two tapered strips with a slanting cut in a thin kerf saw, say a bandsaw. I tried it and it is a pretty hard cut to make, I wouldn't advise it.

But there is another way to do it which would be to thickness the whole 22x150 plank down to 12mm thick and then saw it into two strips 12x72. Then split each strip with a slanting cut to get two pieces 12x20 at one end and 12x50 at the other. In other words get 4 strips (two oars) from each plank.

So in total for 4 oars, you'd need two clear planks of cedar 22x150x3100, a sheet of 4mm ply, and say 5.5m (in 4 pieces) of heavy hardwood 100x50 or the equivalent in 50x175. Looking at the current price list from Robbins Timber of Bristol, who specialise in boatbuilding timber, it would come to £214.53 for 4 oars or £53.63 per oar, plus transport. Affordable I would say. Robbins supply good timber but at a price, so if people can get hardwood or cedar more cheaply they could do better than this. I would say they are nearly the only supplier of decent quality 4mm marine ply, and there is a lot of very poor 4mm ply sold in builders merchants.