BritBike Forum logo
BritBike SponsorBritBike SponsorBritBike SponsorBritBike SponsorBritBike SponsorBritBike SponsorBritBike SponsorThe Bonneville ShopBritBike Sponsor
Upgrade to: Premium Membership | Premium Life Membership | Vendor Membership | Site Sponsor Membership
ShoutChat Box
Comment Guidelines: Do post respectful and insightful comments.
Buy BritBike staff a coffee
Buy BritBike's staff a coffeeStill here since 1996 serving BritBike enthusiasts..
Search eBay for motorcycle parts in following countries
Australia, Canada, France, Holland, Italy, United Kingdom, USA
Random Gallery photo
Member Spotlight
Rick in Seattle
Rick in Seattle
Mercer Island, WA / Austin, TX
Posts: 103
Joined: September 2008
Show All Member Profiles 
Newest Members
Superspin1, Spot, longstroke, Lovatomi, Gary Hawthorn
10898 Registered Users
Top Posters(30 Days)
franko 115
DavidP 73
quinten 68
Popular Topics(Views)
1,046,115 mail-order LSR
a word from..
Forum Statistics
Forums34
Topics68,011
Posts682,486
Members10,898
Most Online14,755
May 5th, 2019
Who's Online Now
69 registered members (998John), 475 guests, and 856 spiders.
Key: Admin, Global Mod, Mod
Previous Thread
Next Thread
Print Thread
Rate Thread
Page 16 of 83 1 2 14 15 16 17 18 82 83
Re: 1928 Ariel Model C [Re: Magnetoman] #716502 11/24/17 7:38 pm
Joined: Aug 2001
Posts: 4,724
gavin eisler Online Content
BritBike Forum member
Online Content
BritBike Forum member
Joined: Aug 2001
Posts: 4,724
This site, http://www.auto-cycle.co.uk/parts.htm , lists " Unsorted box of Ariel pistons " in their inventory. You may have looked there already, thought I would mention it just in case.
I have a Goldstar riding buddy who gets stuff from them ( they have Eddie Dow's old stock). He rates them highly.
They wont be open till Monday but they do answer E mails in shop hours. You might get lucky?, I noticed the Alpha bearings site also lists a few Ariel single pistons but none as old as your model. Fingers crossed , a plus 40 might turn up.
They have NOS home gas tanks for my 71 , at a fair price , which I was delighted to find.


71 Devimead A65 750
56 Norbsa 68 Longstroke A65
Cagiva Raptor 650
MZ TS 250
The poster formerly known as Pod
Support Your #1 BritBike Forum!

Check out British motorcycles for sale:
British Motorcycles on e-Bay UK
British motorcycles on e-Bay North America
Re: 1928 Ariel Model C [Re: Magnetoman] #716530 11/25/17 12:41 am
Joined: Nov 2011
Posts: 5,468
Magnetoman Online Content OP
BritBike Forum member
OP Online Content
BritBike Forum member
Joined: Nov 2011
Posts: 5,468
Originally Posted by RPM
Have a sleeve made.
With a sleeve I could use an original non-cam-ground two-ring piston (see below for clearance discussion). That has its pros and cons. The physical act of sleeving has risks, as does shipping things back and forth. The sprocket I ordered ten days ago from McMaster-Carr to modify for the gearbox won't be delivered until next Tuesday, if then. UPS decided it needed to be shipped all the way to Atlanta and then tour the South before giving it to me, with the last recorded sighting in Mississippi. I have to regard parts like the barrel, head, etc. as irreplaceable since if they go missing there might not be time to find a substitute (although, see below).

Originally Posted by gavin eisler
A pretty substantial package of Gold Star parts from them arrived today. I didn't think to ask about old Ariel pistons. However, thanks to having the weight of an original piston, and two +60 aftermarket cam-ground pistons with oil rings in hand, I think I'm going to go in that direction.

I remeasured the later, cam-ground +30 Hepolite piston that came in my engine. Side-to-side, the dimension that will have the greatest thermal expansion, it is 3.2460", roughly the same as the diameter of the cylindrical "Canadian" +30 piston (3.2450-3.2453"). However, for the front-to-back thrust faces it is 9.7 thou. larger at 3.2557". Since the present bore of the cylinder is 3.2604" at the top the clearance (front-to-back) was 4.7 thou. with the piston that came in the engine.

This raises the question of how much clearance should be used on a non-cam-ground piston? I looked in my 17th, 1935, edition of Dyke's Automobile and Gasoline Engine Encyclopedia and they call for only 0.003" - 0.0035" for Al pistons in this bore size. While that's for water-cooled engines, air cooled won't be that much larger. However, luckily it looks like I won't be playing that particular version of the 'Seize Your Piston' game, instead opting for modernity.

Today I disassembled the bottom end, which meant first removing the timing case. Lacking anything nearly as detailed or accurate as even a Haynes manual, I proceeded slowly and took lots of photos. After removing the sprocket that drives the magneto chain, six screws obviously hold the timing cover in place. The problem in these situations always is, is there a seventh non-obvious screw? Although it didn't budge, after convincing myself there couldn't be any other fasteners holding it, I used the same tool I had made for this purpose for Gold Stars, based on an adjustable-height leveling screw, to carefully press the cover hard enough to get it to move, at which point it came off with a little wiggling.

There's a '.' stamped between teeth in the cam gear with that space directly over a tooth in the crank pinion when the engine is at TDC on the compression stroke. According to Ariel literature (which might be for '29) there should be a corresponding '.' in the pinion, but there isn't. The pinion has three slots broached in it, one directly under the tooth that's under the '.' in the cam gear, with the other two slots offset from being directly under teeth. I marked the pinion with paint and then stamped a '.' after I removed it using a 3/4" A/F socket on the left-hand nut. Funny how a variety of fasteners find their way onto a bike after 90 years, isn't it?

The cam spindle has quite a bit of pitting and scoring in it, as do the cam base circles. There's no pitting in the cam lobes which implies they've been reground. Evidence of that grinding is present. I'll measure the roughness and then polish them with an appropriate grit

The drive side bearing, marked 'Ace' RMS8-2RS, is 2.5" OD, 1" ID and 3/4" thick. It slips in and out of the case easily (too easily...). Its OD is 2.5000" but I haven't yet measured the ID of the housing. In any case, I'll try to find a name-brand bearing to replace it with and deal with the clearance issue after I have it.

Once I removed the crankshaft I found a 7/8" shaft on the timing side and a 1" shaft on the drive side so I decided the easiest way to deal with this for balancing is to buy two bearings with matching ODs. I could have made bushings myself but it would have taken quite a while to make them to the necessary precision (the bores would have to be precisely Concentric with the ODs). As far as I can tell it appears only an original pair of balancing holes are in the flywheels, which means I will be able to determine the original balance factor.

Although I'll wait for the bearings to arrive before making precision measurements, I did take the time to mount the crank between centers in my lathe. Assuming the center drillings are good (and they do look good) the TIR of the drive bearing surface is 0.004" and of the timing bushing surface is 0.0055". The wobble of the drive-side flywheel is 0.015" and of the timing-side flywheel is 0.007". If these values hold up when I make better measurements, they're pretty unimpressive.

Given the condition of the surfaces on the cam, even had I not already ordered a new crankpin I would do so now, before disassembling the crank. Tomorrow I'll polish the cam, inspect (and probably polish) the two shafts on the crank, and then measure clearances of the timing side bushings to see if I need to make new ones.

An interesting discovery I made today is that a BSA B31 (probably a ZB31) cylinder I have on the shelf is a perfect fit on the Ariel bottom end. The cylinder is 1/8" shorter than that of the Ariel but that could be easily dealt with using a compression plate. Conveniently, several months ago I bought a 0.093" Ariel compression plate from Copper Gaskets Unlimited so they already have the pattern for making another one from 0.030" if for some reason they couldn't make a full 1/8". The 71 mm bore of a B31 cylinder can't be taken out to the 88 mm of a 500 cc BSA, but the 7 mm longer stroke of the Arial means it wouldn't have to be. The spacing between bolt holes at the top is ~0.2" different, but that could be solved by plugging the existing holes and tapping new ones. Also, the cutout for the pushrods is slightly narrower but that would be easy to enlarge. Anyway, should the Ariel cylinder fall and shatter this means that it would be a major headache, but not necessarily a disaster.

Offline someone asked me about the precision to which I quote measurements, wondering if I really did mean it.

If I quote a dimension to, say, 0.0001" it means I've checked the relevant micrometer or dial indicator with a calibration rod, ring gauge, or gauge block set (depending on the instrument) before making the measurement so it really does mean 0.0001". That said, to conserve energy in the summer I set the thermostat in my garage to 78 oF when I'm working in it (88 when I'm not) and in the winter to 68 (50 when I'm not). This means the instruments and parts usually haven't equilibrated to the "official" 68 oF (20 oC) for 24 hours for a "true" measurement so there will be some small amount of differential thermal contraction between the steel of the micrometer and the Al (or whatever) of the part that depends on the size of the part. Any resulting error in the quoted dimension is nearly always smaller than anything relevant for motorcycle work, but where it might be relevant I'll mention it, and in cases where it really does matter I will have done a proper measurement at 68 oF.

In the case of weights, such as that of the piston, I have three electronic scales covering different ranges that I calibrate with weights of accuracy better than the resolutions of the scales. Similar comments for voltage, resistance, etc.


Attached Files IMG_6245.JPG
Last edited by Magnetoman; 02/14/18 7:54 pm. Reason: added photo
Re: 1928 Ariel Model C [Re: Magnetoman] #716549 11/25/17 9:13 am
Joined: Oct 2017
Posts: 267
G
George Kaplan Offline
BritBike Forum member
Offline
BritBike Forum member
G
Joined: Oct 2017
Posts: 267
Hello MM, great progress on the Ariel. Re the comments on the accuracy of measurements I seem to remember that you covered the topic previously when you wrote the guide to magneto rebuilds elsewhere on this forum so if anyone is interested in how you do things and what level of accuracy that you aim for then that thread is not a bad start.

I do realise that previous threads are afflicted by the Photobucket hot linking problem but if people cant see your pictures then they should download the patch for either Firefox or Chrome and all will be revealed so to speak.

The comments on the BSA cylinder being close to your Ariel are interesting. Perhaps there is another cylinder from another bike that is even closer if the Ariel cylinder doesn't work out or you feel like you need a spare? I have heard of people substituting cylinders on other bikes in the past so it is certainly a possibility. There must be enough willing old bike people on here to get some dimensions from other cylinders pretty easily. That said a suitable plus 0.040" piston would seem like the best solution.

Keep up the good work I, like a few others on here, will be vicariously preparing for and riding the Cannonball with you.

John

Re: 1928 Ariel Model C [Re: Magnetoman] #716580 11/25/17 4:40 pm
Joined: Nov 2011
Posts: 5,468
Magnetoman Online Content OP
BritBike Forum member
OP Online Content
BritBike Forum member
Joined: Nov 2011
Posts: 5,468
Originally Posted by George Kaplan
Re the comments on the accuracy of measurements I seem to remember that you covered the topic previously when you wrote the guide to magneto rebuilds elsewhere on this forum ...
I'm sure that not for the last time I'll forget and repeat information I've typed elsewhere (perhaps even in the previous post in the same thread...).

Originally Posted by George Kaplan
Perhaps there is another cylinder from another bike that is even closer if the Ariel cylinder doesn't work out or you feel like you need a spare
I don't know what prompted me to put that BSA cylinder on the Ariel yesterday. Probably because I'm interested in the similarities between this engine and Val Page's design of the BSA engine a decade later from which the Gold Star developed so that is always in the back of my mind. As homage to the lineage, and thanks to Kurt Fischer, the Ariel will be wearing a Chronometric listed as having been used on 1952 'Road Model' Gold Stars. Anyway, I wasn't thinking in terms of a spare cylinder when I placed it on the engine. Although, come to think of it, a B34GS cylinder and DBD head could give the bottom end of the Ariel some serious h.p. to recon with...

Speaking of the bottom end, given the loose fit of the ball bearing in the drive side case it can't be counted on to locate the crank. Absent a circlip or other locking mechanism even a press fit in the case wouldn't be sufficient, so what Val Page used for this purpose is a fairly large OD (~1.3") bush on the timing side against which mates a machined surface on the crank. That's another surface I'll need to polish.

The oil pump causes me some concern. It is a simple plunger pushed down by a cam on the camshaft, and pushed back up again by an internal spring. Relying on only a spring to operate it in one direction means that if the plunger binds only slightly in the housing the meager oil flow will cease.

The other aspect of oiling that I'll delve into today are the two "reed valves" that bring oil mist from the crankcase into the timing chest. Both seem to be fine but I'll take them apart to be sure there are no hidden issues. They are based on thin pieces of metal trapped under caps that are press fit into the case so I'll have to figure out how to un-press-fit them, which might not be easy since they are fairly small.

Re: 1928 Ariel Model C [Re: Magnetoman] #716598 11/25/17 6:57 pm
Joined: Dec 2004
Posts: 10,112
kommando Online Content
BritBike Forum member
Online Content
BritBike Forum member
Joined: Dec 2004
Posts: 10,112
Quote
I do realise that previous threads are afflicted by the Photobucket hot linking problem but if people cant see your pictures then they should download the patch for either Firefox or Chrome and all will be revealed so to speak.


If only it was so simple, the fix is currently not working, so we await the next version which will no doubt be got at by PB after a few weeks.

Ah ha,

These are the latest fixes as the original does not work now.

https://chrome.google.com/webstore/...x/kegnjbncdcliihbemealioapbifiaedg?hl=en

and

https://chrome.google.com/webstore/...g/ogipgokcopooepeipngiikdkpmcpkaon?hl=en

I now have all 3 in place and seeing the PB images again.

Last edited by kommando; 11/25/17 7:27 pm. Reason: added new fixes
Re: 1928 Ariel Model C [Re: kommando] #716607 11/25/17 7:28 pm
Joined: Oct 2017
Posts: 267
G
George Kaplan Offline
BritBike Forum member
Offline
BritBike Forum member
G
Joined: Oct 2017
Posts: 267
Originally Posted by Magnetoman
[quote=George Kaplan] The oil pump causes me some concern. It is a simple plunger pushed down by a cam on the camshaft, and pushed back up again by an internal spring. Relying on only a spring to operate it in one direction means that if the plunger binds only slightly in the housing the meager oil flow will cease.


Depending on your level of concern you could fit either an oil pressure gauge or an oil pressure indicator "button" similar to the 1950's Triumphs and, I believe, a few other makes. However adding this might not be simple (I am not familiar with the Black Ariels so not sure if there is a suitable pressure take off point) and adding a gauge or indicator would obviously not conform to the KISS principle.

However the return spring that you describe does not seem ideal. However how often does an oil pump bind? I have no idea if any other bikes used such an arrangement. If no other Black Ariel owners have reported problems then maybe its not such an issue?

Originally Posted by kommando
[quote]If only it was so simple, the fix is currently not working, so we await the next version which will no doubt be got at by PB after a few weeks.


I have Firefox v54 (64 bit) with the latest add-in and I have just checked various threads that I know had the Photobucket problem and I can see all of the pictures.


John

Re: 1928 Ariel Model C [Re: Magnetoman] #716651 11/26/17 1:52 am
Joined: Nov 2011
Posts: 5,468
Magnetoman Online Content OP
BritBike Forum member
OP Online Content
BritBike Forum member
Joined: Nov 2011
Posts: 5,468
Originally Posted by George Kaplan
Depending on your level of concern you could fit either an oil pressure gauge or an oil pressure indicator "button
It would be possible to tap into the gallery but it would not look pretty and I'm not sure how much good it would do unless the gauge was on the handlebars where I could glance at it frequently. However, in a real sense I already have a gauge, in the form of the oil drips in the sight glass on the tank. If the drips stop dripping, the pump has stopped pumping (or I forgot to fill the tank -- either way, the same effect on the engine). Unfortunately, since there's no realistic way to re-engineer the oil pump I'll have to live with its design and hope it doesn't bind.

I made a lot of progress today, some of which is visible. The cases were pretty clean but I soaked them in my new-favorite mix of diesel and Gunk anyway and spent some time removing old, hard joining compound remnants. There are two small holes that allow oil to get to the bushes and they could easily be blocked by a wayward chip of debris.

Measured roughnesses (Ra in microinches) of the various shafts ranged from ~24-30 for the timing side of the crank, to ~125-135 for the outside shaft of the camshaft due to pitting. Ideally, all would be less than 10. In the case of the camshaft the roughness value is deceptive in that what matters most is the roughness on the plateaus between pits since that will be the bearing surface. The pits represent absence of material, i.e. valleys between plateaus rather than between peaks.

All three bushes in the timing side have clearances ~0.003-0.004", whereas they should be no more than 0.001", so they'll have to be replaced. If I can't buy them I'll have to make them.

There are no locating dowels in the crankcase halves so it is possible to assemble them with a slight twist due to the clearance holes for the 5/16" and 3/8" studs. This would result in small steps at the front and back of the crankcase mouth. So, I faced a piece of 4"x4"x1/2" scrap Al and drilled the necessary four holes spaced 3-3/8" to use when assembling the two halves. After doing this I confirmed on a surface plate that the machined surface was still flat.

Starting with a 1"-dia. rod for the drive side I machined one end of it 7/8"-dia. for the timing side. I then bolted the cases together with this rod protruding from either end, using the flat plate I had just made to ensure the cases went together with no twist. When done I put the cases on the surface place, shimmed so the rod was level, and then measured the side-to-side tilt of the crankcase mouth.

It's no problem if a cylinder tilts slightly fore or aft but if it's tilted sideways the piston has to slide back and forth on the gudgeon pin to follow that tilt on each revolution. The crankcase mouth isn't perfectly smooth but it was no problem to determine that it tilts with respect to the crankshaft axis by ~0.045"-0.050". Since the stroke is ~3.7" this means the piston has to slide side-to-side by approximately the same ~0.045" twice on every revolution. Knowing this I'll now remove the studs and followers from the case and mount the crankcase on the mill indexed to the crankshaft axis in order to face the mouth.

Since the crankshaft was out of the engine I oriented it so the rod was horizontal and measured the weight of the small end. I won't be able to do this properly until the crank is apart, but this figure will be close enough for present purposes. Adding that 264 grams to the weight of the complete piston assembly that was in my engine when I bought it I get 730 grams. Doing the same for the original piston that came in the engine 90 years ago results in 769 grams. For the purposes of making an estimate, assume the final balance factor comes out at 66%. This difference of 39 grams in the weights of the two pistons means a difference of 26 grams at the flywheel from the value Ariel decided it "should" have when they determined the balance factor. Plugging 26 grams into a calculation, this weight rotating at a diameter of 8-1/4" results in an additional oscillating force of 168 lbs. at 5000 rpm beyond that which it would have had. If the engine is at 3000 rpm it's still an additional 60 lbs. hammering at the frame and the rider. This is why I'm been spending the effort required to "properly" balance this engine with whatever piston I end up using.

Re: 1928 Ariel Model C [Re: Magnetoman] #716698 11/26/17 1:42 pm
Joined: Feb 2014
Posts: 582
C
chaterlea25 Offline
BritBike Forum member
Offline
BritBike Forum member
C
Joined: Feb 2014
Posts: 582
Hi MM,
Earlier you announced that the drive side main was "loose" in the case
I would "fix" that problem first, then check that the timing side bush is Concentric with the case bore,
I would align the cases as best possible and then ream two of the bolt holes,
Then make dowels to suit them so reassembly accuracy can be repeated
Correct any misalignment across the main bearing housings, Fit drive side bearing, then line bore the timing side bush to it
I would do all this before correcting the cylinder face
I would think that 0.001in. is too tight for the cam bushes, this is a total loss lubed engine you need "space" for the oil to get into the bush bores

John


Re: 1928 Ariel Model C [Re: Magnetoman] #716753 11/26/17 6:36 pm
Joined: Nov 2011
Posts: 5,468
Magnetoman Online Content OP
BritBike Forum member
OP Online Content
BritBike Forum member
Joined: Nov 2011
Posts: 5,468
Originally Posted by chaterlea25
... the drive side main was "loose" in the case
I would "fix" that problem first, then check that the timing side bush is Concentric
Correct any misalignment across the main bearing housings, Fit drive side bearing, then line bore the timing side bush to it
I would do all this before correcting the cylinder face
Thanks for mentioning this. Even though I said I would "now" face the mouth I meant now it would get added to my list of things to do. I'm deliberately taking this slowly to try to avoid making any permanent changes like this in the wrong order. One reason I write in some detail about what I'm doing is in the hopes someone like you catches any mistakes before I make them. Specifically:
Originally Posted by chaterlea25
I would think that 0.001in. is too tight for the cam bushes, this is a total loss lubed engine you need "space" for the oil to get into the bush bores
Hmm. Your statement has implications for the timing side crankshaft bush as well.

First, for what it's worth, despite the timing side bush being 1-3/4" long the faux crankshaft I made rotates in the cases without problem so right now the alignment is pretty good. The ~0.004" clearance (0.08741"-0.8745" crank & 0.08782" bush) helps allow this, but I suspect it means the original machining of the cases is responsible for the good alignment. That, plus an aftermarket bush having reasonably Concentric ID and OD, which would be expected. But, back to the clearance issue.

Reading the Ariel 'Owners Guide' is one thing, but there's no substitute for tracing the oil system myself in order to understand it. What the section 'Engine Automatic Lubrication Explained' says is "Oil is supplied by means of an internal mechanical pump... delivers it to a recess in the flywheel, from which it travels to the big end bearing via a hole drilled through the flywheel." What I now know this means is while it appears a drilling delivers pressurized oil to the timing side crankshaft bush, what actually happens is it delivers it to a groove beside the bush from where it travels to an annular volume on the face of the flywheel defined by a machined surface in contact with the face of the bush and a ring on the flywheel that overlaps a circular projection on the crankcase. Inside that annular volume is a hole drilled at an angle from which centripetal force takes the oil to the big end. After leaving the big end it joins the oil-rich smog inside the crankcase, some of which condenses and makes its way to one hole at the top of one end of the timing side bush.

The crankcase smog gets pumped into the timing case through two one-way reed valves where it condenses and fills the case to a level above another hole that leads to the timing side bush. The level of the liquid oil is maintained by a 0.08" hole back into the crankcase. Interestingly, this hole could have been drilled directly over the hole leading to the timing side bush to drip liquid onto it, but it wasn't. This liquid timing case oil also gets splashed around by the cam gear and pinion where some of it enters holes at the top of the outer camshaft bush. The inner camshaft bush has no hole over it so it relies completely on whatever oil somehow enters it via either end of the shaft (none of the shafts have scrolls on them to encourage oil flow). Because having actual liquid oil in the timing case is essential, when the case is drained every 1000 miles through a plug at the bottom, it is necessary to refill it with 1/4 pint rather than waiting for that to happen by condensation.

OK, back to clearances. None of the three bushes is pressure fed, which is the same as for the bushes in a BSA gearbox. I use that as an example because my measurements of a few layshafts on the shelf indicate that BSA made their 11/16" shafts to allow for 0.002" clearance. Given this, if you were me, would you make new camshaft bushes given their 0.003" clearance? Although the Ariel's crankshaft is larger (7/8") it appears its clearance of 0.0037-0.0041" gives me no choice, although I'll remeasure everything before replacing anything.

Originally Posted by chaterlea25
I would align the cases as best possible and then ream two of the bolt holes,
Then make dowels to suit them so reassembly accuracy can be repeated
That's not a bad idea, but there are pros and cons to it. The pro is it would make it easier for some future rebuilder to accurately assemble the cases even if they didn't realize they needed to be concerned with this. The con is it would require metal to be removed from the cases to make room for the dowels, and I try very hard not to remove metal from original pieces whenever possible. The plate I made allows me to accurately assemble the cases myself as many times as needed, and I will have it along on the Cannonball if, god forbid, the engine has to come apart at all, let alone that far apart.


Re: 1928 Ariel Model C [Re: Magnetoman] #716767 11/26/17 7:19 pm
Joined: Dec 2004
Posts: 10,112
kommando Online Content
BritBike Forum member
Online Content
BritBike Forum member
Joined: Dec 2004
Posts: 10,112
I can understand the logic behind giving a bigger clearance when there is not much oil about in an effort to get it in there but the greater the clearance the quicker it will also exit so do not overdo it.

From

http://edge.rit.edu/edge/P14453/public/Research/2-_LEADER_-_Understanding_Journal_Bearings.pdf

6. Stay within design guidelines on clearances. General rule is 1.5 mils (0.0015 inches) per inch of shaft journal diameter. 3.0 mils/inch diameter is
excessive clearance in most cases.

Also table on page 8 for Hydrodynamic bushes

http://www.leonardocentre.co.uk/Media/Default/Documents/Book_1_bearings.pdf

which confirms the figure but relates it to speed of shaft and journal size in a more detailed manner.

Re: 1928 Ariel Model C [Re: RPM] #716770 11/26/17 7:26 pm
Joined: Sep 2007
Posts: 1,130
gREgg-K Offline
BritBike Forum member
Offline
BritBike Forum member
Joined: Sep 2007
Posts: 1,130
Thanks RPM, for the pointer to Ferrea. I'll give them a try.

. Gregg


Spyder Integrated Technologies
Lucas, BTH, & Miller Magneto & Dynamo Restoration
SMITHS Chronometric Restoration
[email protected]
Re: 1928 Ariel Model C [Re: Magnetoman] #716820 11/27/17 12:50 am
Joined: Nov 2011
Posts: 5,468
Magnetoman Online Content OP
BritBike Forum member
OP Online Content
BritBike Forum member
Joined: Nov 2011
Posts: 5,468
Kommando, thanks for the links. Life would be simpler if I would just buy bearings and bushes on blind faith from dealers, install them, and blame myself if/when they failed after some number of miles. But, nooo, it isn't simple because I know details like 'C' clearances on bearings, properties of different bronzes, boundary layer lubrication, line boring, proper alignment of the deck, etc., etc. So, instead of blissful ignorance I have to spend hours doing research.

For those of you who are still leading a blissful life, in this post you can see with your own eyes (or, you used to, before Photobucket stepped in) a bearing that was nonfunctional, with the problem due to it having been mis-machined by ~1/64". I had ordered two and both were mis-made the same way. Despite this the supplier responded:

We sell approximately 80 each year and we are not aware of any concerns raised.

Even if a supplier stocked camshaft bushings for 1928 Ariels the only way I would know that, in turn, their suppliers had manufactured them using a proper composition of bronze (instead of one of scores of compositions that look the same but that have the wrong properties) that would last 4000 miles is if they actually lasted 4000 miles, rather than wearing out in 1000.

Moving on from bushes to bearings, relevant background information is I have found sources for unshielded RMS8 bearings made by FAG and SKF, both unmarked which means they would be of CN ("normal") clearance. As an aside, although the bearing now in my Ariel has a shield, presumably the original ones did not. Because of the design all but an annular area of width 0.028" of the oil's escape route through the bearing is blocked by the case so I have to wonder if the inner races of 1928 bearings might have been just that much wider. In any case, I'll either try to remove the shield from the current bearing and transfer it to whatever name-brand bearing I buy or, if that doesn't work, make my own shield from shim stock.

The OD of my current bearing is 2.4998" and the bore in the crankcase is slightly larger and slightly oval, with clearance varying between 0.0007" and 0.0017". The sprocket cush drive fastens securely to the bearing's inner race via a spacer and Ariel calls for an end float of 0.008-0.012". What is means is the crankshaft could/would move this bearing back and forth in the case by this amount at present. So, what to do? Three possibilities open for comment are:

-- Accurately register the case and open up the hole just enough to make it round, i.e. leaving it oversize by 0.0017" all the way around. Electroplate 0.0018" on the OD of a new bearing which would result in a 0.0019" interference fit. The risk would be a ~0.002" interference fit would make a CN bearing too tight.

-- There's room to open up the case by ~1/8"-dia. at most so a 1/16" wall thickness steel or brass sleeve tightly pressed in the case (0.0025") could be made that also allowed for a somewhat looser press fit (0.0015') for the bearing. The risks with this approach would be overdoing the sleeve press fit and cracking the case, and that a 0.0015" bearing press fit could make a CN bearing too tight.

-- Locktite 648 would be the easiest of the three possibilities. OK, OK, Loctite doesn't have the best reputation for main bearings, but Ariel didn't deign to include any kind of mechanical retention system so they expected (hoped for, crossed their fingers that,...) a press fit would suffice. According to tables and a formula in 'Machinery's Handbook' it takes 1.1 ton for a 0.002 interference fit of 2.5"-dia. steel in cast iron, so the force would be less in Al. If you study the Loctite 458 data sheet you'll see that fully cured on properly prepared Al the sheer strength is 2900 psi which, for the surface area of the bearing, means it would take 8.5 tons to move it.




Re: 1928 Ariel Model C [Re: Magnetoman] #716853 11/27/17 9:19 am
Joined: Dec 2004
Posts: 10,112
kommando Online Content
BritBike Forum member
Online Content
BritBike Forum member
Joined: Dec 2004
Posts: 10,112
Loctite is definitely an issue when used to locate steel ball bearings in an alloy case, tried it twice and it took less than 1000 miles both times before failing, my theory is that the differential expansion rates aluminium to steel break the bond. If my theory is correct then it will not fail in the same way on a steel ball bearing in a cast iron case, but as it's only a theory then you best stick with either a sleeve or electroplate the bearing. Both are tried and tested leaving you only with the CN C3 decision, there is an overlap between the 2 tolerances.

Re: 1928 Ariel Model C [Re: kommando] #716861 11/27/17 11:29 am
Joined: Oct 2001
Posts: 1,159
S
Shane in Oz Online Content
BritBike Forum member
Online Content
BritBike Forum member
S
Joined: Oct 2001
Posts: 1,159
It looks like Loctite Bearing Lock is out.

Rather than plating the outer race or boring and sleeving the crankcase, would it be practical to use a metric bearing with a 64mm o.d.?
For example, the AB44203S01 is 26 x 64 x 16 and 28BC06S10 is 28 x 64 x 15. It's generally preferable to sleeve a steel shaft than an alloy housing. Those particular bearing are 3mm and 4mm narrower, which would have an effect on load rating, but may possibly allow room to machine a circlip groove if one wishes hard enough.

Re: 1928 Ariel Model C [Re: Magnetoman] #716878 11/27/17 2:33 pm
Joined: Aug 2001
Posts: 4,724
gavin eisler Online Content
BritBike Forum member
Online Content
BritBike Forum member
Joined: Aug 2001
Posts: 4,724
I tried to take up play on my A65 which had an oval mainshaft, play between DS main and bearing inner, used bearing fit loctite,didnt work.
I did successfully repair a cracked marine engine, the main bearing housing had cracked, the iron cases were welded up then the bearing housing enlarged to take a top hat sleeve, then bored this sleeve to suit, there was a lot of meat to play with and it worked out well. The top hat sleeve was secured by c/ sunk screws around the flange. A bit like some Norton motors i have seen. Although with only a 1/16th to play with , hmmm.
Plating or grinding down an OS bearing seems like a better fix.


71 Devimead A65 750
56 Norbsa 68 Longstroke A65
Cagiva Raptor 650
MZ TS 250
The poster formerly known as Pod
Re: 1928 Ariel Model C [Re: Magnetoman] #716884 11/27/17 3:13 pm
Joined: Aug 2006
Posts: 530
RPM Online Content
BritBike Forum member
Online Content
BritBike Forum member
Joined: Aug 2006
Posts: 530
I have never had any luck with the loctite bearing fit into alloy cases either.

My mentor and friend Ed Mabry ( builder of the first sit on and ride 250 mph motorcycle and SEMA member No. 12) had a sign on his shop wall that stated.

" To much knowledge can often be a limiting factor"

We could also quote the great EJ Potter aka The Michigan Madman

" Ignorance can be a powerful tool when applied at the right time. Often and usually surpassing knowledge."


Just messing with MM. I think your approach is well worth it and I have learned a few things reading this post.



Re: 1928 Ariel Model C [Re: Magnetoman] #716889 11/27/17 3:52 pm
Joined: Nov 2011
Posts: 5,468
Magnetoman Online Content OP
BritBike Forum member
OP Online Content
BritBike Forum member
Joined: Nov 2011
Posts: 5,468
Originally Posted by kommando
Loctite is definitely an issue when used to locate steel ball bearings in an alloy case, tried it twice and it took less than 1000 miles both times before failing
Before pounding a stake through the heart of this possibility, did you note which of the many Loctites you used? Also, did you use an activator and, if so, which one?

Originally Posted by Shane in Oz
Rather than plating the outer race or boring and sleeving the crankcase, would it be practical to use a metric bearing with a 64mm o.d.?
Hmm, I hadn't thought of that. Thanks for raising the possibility. Even though the extra 0.197" diameter of either of those bearings is too much to accommodate by enlarging the cavity your suggestion will send me to my shelf of bearing books to look for alternatives.

Originally Posted by gavin eisler
Plating or grinding down an OS bearing seems like a better fix.
I'll add grinding the OD of a 64 mm race down to 63.6 (2.5039") to the list of possibilities.

As I mentioned in my previous post, the fact the ID of the sprocket side of the cavity almost shields the bearing made me think the original bearing Ariel used might have had different specs. Now add to this the fact the current RMS8 is 3/4" wide but the cavity is a devilish 0.666" deep so the bearing sits proud by 0.084". Instead, if the original bearing had been 5/8" wide it would be recessed by 0.041 (although, not enough for a circlip, I'm afraid). That difference easily could have been accommodated by a 1/8" spacer on the crankshaft. Unfortunately, even if I could find such a bearing it doesn't help me with the retention problem.(*)

Originally Posted by kommando
electroplate the bearing. ...tried and tested leaving you only with the CN C3 decision, there is an overlap between the 2 tolerances.
Even if Locktite would work, and even though it would be easier if it did, more often than not metal provides permanent solutions to problems while glue does not. So, after locating whatever bearing I will end up using, it looks like it will be time to mask the internals and fire up the electroplater.

All of the 'C' grades overlap with each other, e.g. a loose C(i) is tighter than a tight C(i+1). I don't have a choice of grades in the case of the two name-brand RMS8s that I've found but perhaps I will if a mythical 5/8"-wide version turns out to be real. And available.

As an aside, it's possible to measure the actual clearance of a given bearing. According to one table a force of 49 N (11 lbs.) on the outer race of a 2.5" OD bearing rated CN will displace it sideways with respect to the inner race by 5 micrometers (197 microinches). The smallest divisions on the most sensitive mechanical dial indicator I have, a Mahr Supramess, are 0.5 micrometers (as an aside to this aside, it's (barely) possible to control the position of the needle with a finger to ~1 division on this mechanical device, i.e. to the wavelength of blue/green light).

It would be easy to make a base to which the inner race could be rigidly clamped and the outer race would be free to be pushed back and forth against the tip of the indicator. Making such a test rig has been on my to-do list for a few years, ever since I came into a small collection of used Gold Star engine ball and roller bearings. All of them "feel" good, but a simple test rig would allow me to determine their actual clearances.

RPM: I've certainly seen plenty of examples over the years of ignorance being bliss. Unfortunately, I was born cursed with a thirst for knowledge that has haunted me my entire life. Luckily, although curiosity killed the cat, I'm still alive despite several close calls when I was young.


(*) Series RLM bearings are narrower. In particular, RLM9 is 5/8" wide. The SKF catalog rates it for a dynamic load of 3350 lbs. whereas the wider RMS8 is only ~7% higher at 3600 lbs. Update: The RLM9 is for a 1-1/8" shaft, but adding a sleeve to the crankshaft wouldn't be a problem. The RLM8 is for a 1" shaft but its OD is 2-1/4" so it would require a sleeve for the case, which would be more work. If not for the crankpin nut, with the narrower 5/8" bearing it would have been possible to bolt it place.

Last edited by Magnetoman; 11/27/17 6:24 pm. Reason: first added (*) and and hour later an update
Re: 1928 Ariel Model C [Re: Magnetoman] #716912 11/27/17 7:42 pm
Joined: Feb 2014
Posts: 582
C
chaterlea25 Offline
BritBike Forum member
Offline
BritBike Forum member
C
Joined: Feb 2014
Posts: 582
Hi MM, and All
Going back to my earlier observations on cam bush clearance, 0.0015 - 0.002 in. would be my aim
Finding 0.003 is an acceptable value on an old engine, and at 0.004 I would advise replacing
That said,If you read the workshop manual for Indian Chief engines they do not advocate changing bushes till 2 x the above value

As to the main bearing issue, I believe the problem was caused by the crank running out of line ?
A further bearing to consider is a 6305 (25x62x17) modified with an imperial 1" bore ? ( I think that they might be available to order?)
Ok so you would need to remove less from the casing to fit the sleeve, but going that far would it be easier to repair for the original size bearing
I would opt for using the original size bearing if possible as the casing housing is somewhat oval then it will need "repair" anyway
I would opt for the 0.002 interference fit and a C3 bearing


The big end lubrication system from your description sounds just like the JAP engines of the period (same designer I believe ?)
The system relies on the crankcase being airtight to raise a vacuum inside the cases as the piston rises, the breather valves allow pressure to escape into the timing case as the piston descends
To work properly the timing side bearing needs to be shielded to prevent air being drawn in
So either the mainshaft / sleeve needs to be a tightish fit where the shaft comes through the case or fit a bearing with Z or RS on the inside
(similar to what some people do with Gold Star engines)
Cases are often worn due to running with worn main bearings in the past

John









Re: 1928 Ariel Model C [Re: Magnetoman] #716916 11/27/17 8:22 pm
Joined: Dec 2004
Posts: 10,112
kommando Online Content
BritBike Forum member
Online Content
BritBike Forum member
Joined: Dec 2004
Posts: 10,112
It was in a silvery grey tube just called 'bearing fit' but no number, nearest current equivalent is Locitite 660, I did not use an activator. I assumed after the first failure it was my process that was at fault, so second time I was much more meticulous, both times I used acetone to clean all surfaces and allowed the product to set at room temp over first time 1 day and 2 days the second. Both times when it failed the bearing was loose in the housing and sections of the loctite were in the sump.

Tech sheet for 660

https://www.bearingboys.co.uk/uploads/660-EN.pdf

Re: 1928 Ariel Model C [Re: Magnetoman] #716949 11/28/17 10:51 am
Joined: Jan 2014
Posts: 19
K
KevinN Offline
BritBike Forum member
Offline
BritBike Forum member
K
Joined: Jan 2014
Posts: 19
Charles,

All this talk about clearances reminded me that Nicholson's book had an Ariel section in it. Unfortunately the cam bushing is not listed in the clearance table. I thought I'd post the table and a few other things anyway, in case they are useful. This is from a 1965 edition, so I'm not sure how applicable it is to your machine.

[Linked Image]

[Linked Image]

[Linked Image]

[Linked Image]




Kevin


.

Re: 1928 Ariel Model C [Re: Magnetoman] #717017 11/28/17 8:14 pm
Joined: Nov 2011
Posts: 5,468
Magnetoman Online Content OP
BritBike Forum member
OP Online Content
BritBike Forum member
Joined: Nov 2011
Posts: 5,468
Originally Posted by kommando
It was in a silvery grey tube just called 'bearing fit' but no number,
Thanks for this. As you know Loctite comes in at least 127 variations, but all are subject to the same differential expansion issue you mentioned in a previous email.

Originally Posted by KevinN
All this talk about clearances reminded me that Nicholson's book had an Ariel section in it.
Kevin, thanks for posting that. I'm looking for as much technical information as I can find on this machine. I actually have a full set of editions going back to Nicholson's 1st, which I looked through when compiling my 'shop manual'. Unfortunately, even though my Ariel was only 20 years old then, nothing in Nicholson is directly related to it.

Originally Posted by chaterlea25
The big end lubrication system from your description sounds just like the JAP engines of the period (same designer I believe ?)
The system relies on the crankcase being airtight ...
Yes, Val Page came to Ariel after designing JAP engines. The crankcase isn't quite sealed because there's a pipe about 3" above the bottom that takes oil mist into the primary case where it condenses and drips on the chain. However, a rough measurement shows the total area of the breather holes into the timing chest is ~7x that of the breather pipe so only a fraction of the oil mist ends up on the chain. The timing chest also has an outlet pipe for excess oil, taking it to the drive chain.

Originally Posted by chaterlea25
As to the main bearing issue, I believe the problem was caused by the crank running out of line ?
I would opt for using the original size bearing if possible as the casing housing is somewhat oval then it will need "repair" anyway
I would opt for the 0.002 interference fit and a C3 bearing
The big 'if' is if I can find an appropriate C3 bearing. I've posted a question about this on the Ariel Owner's Club website in the hopes some there has useful experience on bearing choices to offer.

The high spots in the housing and the surface of the bearing have been polished by movement of the bearing. This could have happened when the previous rebuilder pressed the bearing into a too-large oval housing or, more likely, the polishing marks and ovalness are due to the bearing having "walked" because the previous rebuilder assembled the crankshaft with the shafts not accurate aligned. Either way, and not for the first or last time, I'll say that I'm glad I'm completely rebuilding this already-"rebuilt" bike.

My measurements showed that to make the hole for the bearing round again requires removing ~0.001" from the diameter. However, to do this requires mounting the mating face of that half of the crankcase precisely parallel with the mill bed in order to make that hole precisely perpendicular. As chaterlea25 knows, machining anything is 90% fixturing and 50% everything else, and repairing the crankcase is an example of that.

The case has 8 machined "ears" on the outside for clamping the halves together and for mounting the engine in the frame. I set the case on two 1/4"-wide parallels on the surface plate that spanned two of the mounting lugs on opposite sides. Since the machined surfaces on the case aren't perfectly smooth I set another parallel on top to average out the roughness and then used a test indicator to run across the top of it. Working my way around the entire case two lugs at a time the maximum tilt was ~0.005". This is great because it means I will be able to mount the case to the mill using those lugs, with shim stock to make final tweaks. I then will find the approximate center of the ovaled hole that will requires the least amount of material to be removed to make it round again.

The actual machining of the hole will take about a minute, but only after at least a half day setting everything up. Toward that end, yesterday I machined two standoffs for holding the case on the mill bed, consisting of 1-1/2"x1-1/2"x4-1/2" blocks with 3/8" through holes to bolt to the table, and 3/8-16 tapped holes for bolting the case to the blocks. When I'm ready to do the machining, shims between the blocks and case will adjust the mating surface to be flat.

1960s catalogs from Hoffmann, SKF, R&M, and FAG all have "single row deep groove ball bearings" of the size of the RMS8 that was in the engine. What's interesting is the wide range of capacities shown for these bearings, all of which would be identical in actual performance:

R&M: 'Basic Capacity' 460 lbs at 4000 rpm
Hoffmann: 'Safe Working Load' 1300 lbs at 4000 rpm
SKF: 'Basic Capacity' static 2200 lbs., dynamic 3250 lbs., Max. rpm 13,000
FAG: 'Load Rating' static 4000 lbs, dynamic 3800

A delayed delivery of bearings needed to mount the crank on my rollers is now scheduled for today. As soon as those bearings actually show up I'll be able to make accurate measurements of the current state of the crankshaft. The big end nuts have socket head cap screws retaining them so it has been rebuilt after leaving the factory.

Attached Files IMG_6228.JPG
Last edited by Magnetoman; 02/14/18 7:46 pm. Reason: added photo
Re: 1928 Ariel Model C [Re: Magnetoman] #717021 11/28/17 8:31 pm
Joined: Aug 2001
Posts: 4,724
gavin eisler Online Content
BritBike Forum member
Online Content
BritBike Forum member
Joined: Aug 2001
Posts: 4,724
Hi MM,
The National Motorcycle Museum in the UK has BMS copies of Ariel period data here.
http://www.nationalmotorcyclemuseum...ith-bms-motorcycle-manuals/ariel/page/2/
Maybe you have these already?


71 Devimead A65 750
56 Norbsa 68 Longstroke A65
Cagiva Raptor 650
MZ TS 250
The poster formerly known as Pod
Re: 1928 Ariel Model C [Re: Magnetoman] #717029 11/28/17 9:52 pm
Joined: Feb 2012
Posts: 189
L
L.A.kevin Offline
BritBike Forum member
Offline
BritBike Forum member
L
Joined: Feb 2012
Posts: 189
MM, I've only recently stumbled on this thread, and though I can't add much to it, I must say this is a really great read. Thanks, and I hope to get some time to be there at the finish to cheer you on your Ariel and my good buddy Richard on his Norton.

Kevin

Re: 1928 Ariel Model C [Re: Magnetoman] #717056 11/29/17 3:19 am
Joined: Nov 2011
Posts: 5,468
Magnetoman Online Content OP
BritBike Forum member
OP Online Content
BritBike Forum member
Joined: Nov 2011
Posts: 5,468
Originally Posted by gavin eisler
Maybe you have these already?
Thanks very much for the link, but paper copies of the factory literature came with the bike. Also, the Ariel Owner's Club has pdfs of several 'how-to' magazine articles from the time.

Originally Posted by L.A.kevin
I hope to get some time to be there at the finish to cheer you on your Ariel and my good buddy Richard on his Norton.
I hope to see you there on your to-be-rebuilt Gold Star.

Re: 1928 Ariel Model C [Re: Magnetoman] #717105 11/29/17 4:03 pm
Joined: Nov 2011
Posts: 5,468
Magnetoman Online Content OP
BritBike Forum member
OP Online Content
BritBike Forum member
Joined: Nov 2011
Posts: 5,468
I've now had it confirmed by someone in the AOMCC that, although no details are in the '28 Ariel catalog, the '27 and '31 catalogs list an MJ1 as the drive-side bearing. I hadn't thought to look in my own copy of the '27 catalog (note to Gavin -- it's a BMS photocopy that came with the bike along with other literature), but indeed it does list an 'M.J.1 Ball Bearing' at a price of 12s 0d.

Now on order is an RHP (successor company to R&M, formed in '69) MJ1 bearing, sealed but of unknown clearance. I'll have to wait until it shows up to see what clearance is stamped on it, but I'd bet it will be a CN. As mentioned in an earlier post, there is an overlap of clearance grades so I'll make a jig to measure the actual clearance then plate its OD for a ~0.0015-0.002" press fit (depending on the measured bearing clearance) in the waiting-to-be-repaired crankcase hole.

The gearbox sprocket with hub I ordered long, long ago finally was delivered by UPS yesterday after its tour of the country, but the bearings I need to balance the crank have been delayed until tomorrow. It's a good thing I'm able to work on things in parallel because linearly, this would take forever. Still, every item that (eventually) arrives brings me a step closer to the finish.

Page 16 of 83 1 2 14 15 16 17 18 82 83

Home | Sponsors | Newsletter | Regalia | Calendar | Bike Project | BritBike Museum | Spiders Cartoons | DVD- Manuals & Parts books
Upgrade to: Premium Membership | Premium Life Membership | Vendor Membership | Site Sponsor Membership
Powered by UBB.threads™ PHP Forum Software 7.7.3