Britbike forum

Classic British SparesKlempf British PartsBaxter CycleBritBike Sponsor SteadfastCyclesThe Bonneville ShopLowbrow CustomsSRM EngineeringGirling Classic MotorcycleLucas Classic MotorcycleHepolite PistonsIndustrial tec supply

Upgrade your membership: Premium Membership Gold Membership


New Sponsor post
Cyber Monday Sale - 10% off at The Bonneville Shop
by The Bonneville Shop - 11/29/21 5:34 pm
New FAQ post
How to find my own threads?
by reverb - 11/20/21 3:55 pm
Manuals on DVD - Buy 4 for 3
All 4 DVD Manual
Member Spotlight
kommando
kommando
Scotland
Posts: 11,449
Joined: December 2004
ShoutChat
Comment Guidelines: Do post respectful and insightful comments. Don't flame, hate, spam.
Top Posters(30 Days)
NickL 79
Rohan 77
Top Likes Received (30 Days)
kevin 22
Newest Members
old ion limey, muzz750, Motorcycles and Art, SteveDee, Gus_Carrero
11,924 Registered Users
Who's Online Now
5 members (DJinCA, Robbo6, BrizzoBrit, Mark Z, kevin), 21 guests, and 15 robots.
Key: Admin, Global Mod, Mod
Previous Thread
Next Thread
Print Thread
Page 4 of 6 1 2 3 4 5 6
Joined: Nov 2011
Posts: 7,493
Likes: 528
Britbike forum member
OP Online Content
Britbike forum member
Joined: Nov 2011
Posts: 7,493
Likes: 528
Originally Posted by Chris the camper.
'If you don't measure you don't know.'
Referring to the data in my previous post, imagine a correctly-sized '500' was a little too lean, but the '510' you substituted actually flowed 3½ sizes more than its marked size, at 545. Now the bike is much too rich, so you try a '490' out of desperation, but the one you picked from the box happened to flow 3½ sizes less than its marked size, at 455, so it's so lean that you put a hole in your piston. At that point you would abandon your Bonneville record attempt out of frustration.

Originally Posted by kevin
the graph shows cfpm of air through the different jets. how are you calibrating air to fuel?
unless those actual jets were previously selected to be best fits for that labelled fuel flow?
The page of calculations I left out says the air flow over the entire size regime of AMAL jets should be linear, and the first plot confirms that theoretical prediction. For the graph I plotted my measurements on the Y axis, and took AMAL's measurements (i.e. the numbers stamped on the jets) at face value for the fuel-flow rates and plotted them on the X-axis. Of course, like any experiment, data is scattered above and below the average value. However, the experiment is good enough, i.e. the scatter is small enough, that we can conclude with a high degree of confidence that the air flow rate is linearly proportional to the fuel flow rate, as predicted by theory.

Irrespective of the rest of the graph, if every '500' jet actually flowed the same amount of fuel, as they should if the markings were correct, the air flow measurements of all those jets would have been the same.

1 member likes this: Chris the camper.
British motorcycles on eBay
Joined: Nov 2021
Posts: 69
Likes: 3
C
Britbike forum member
Offline
Britbike forum member
C
Joined: Nov 2021
Posts: 69
Likes: 3
I set both carbs on my Commando going by the AFR readings of an O2 sensor plumbed into one of the exhaust pipes. I'll need to compare the pilot, needle and main jets on the 'untested' carb to the tested carb.

Joined: Nov 2011
Posts: 7,493
Likes: 528
Britbike forum member
OP Online Content
Britbike forum member
Joined: Nov 2011
Posts: 7,493
Likes: 528
Originally Posted by Chris the camper.
I set both carbs on my Commando going by the AFR readings of an O2 sensor plumbed into one of the exhaust pipes.
Mechanical measurements aren't the only ones that interest me, with some results from my Mark I AFR system here and my upgraded Mark II system shown here.

Personally, though, I find it significantly easier to get the carburetion perfectly balanced on bikes with only one carburetor...

Joined: Dec 2013
Posts: 5,809
Likes: 141
DOPE
Online Confused
DOPE
Joined: Dec 2013
Posts: 5,809
Likes: 141
its interesting that there really does appear to be a change in slope at exactly 1000. undocumented laboratory quirks.

bonneville is problematic because you have no convenient means of testing mixture there. the number of runs you get is limited, the elevation is much higher than most places at which you will have done your testing, and temperature and relative air density are not particularly consistent, aiui.

you have to wing it on mixture using best guesses on what you think your jets will provide. your suggestion to pre-test your jets for flow is good advice.


i'm old enough to remember when patriotism meant not trying to overthrow the government.
Joined: Nov 2011
Posts: 7,493
Likes: 528
Britbike forum member
OP Online Content
Britbike forum member
Joined: Nov 2011
Posts: 7,493
Likes: 528
Originally Posted by kevin
bonneville is problematic because you have no convenient means of testing mixture there. the number of runs you get is limited, the elevation is much higher than most places at which you will have done your testing, and temperature and relative air density are not particularly consistent,
Having been in a car within a few miles of the salt flats when I was a kid, and having driven through Salt Lake City just a few years ago, I write with great authority about tuning bikes at Bonneville...

Joined: Oct 2001
Posts: 1,895
Likes: 216
Britbike forum member
Offline
Britbike forum member
Joined: Oct 2001
Posts: 1,895
Likes: 216
Originally Posted by Magnetoman
Having been in a car within a few miles of the salt flats when I was a kid, and having driven through Salt Lake City just a few years ago, I write with great authority about tuning bikes at Bonneville...
As Tom Lehrer would say, be prepared.
That will come in very handy when The World's Fastest Ariel and the lesser marque find their way to Bonneville. Don't forget to pack the budgie smugglers and sand shoes.

Sorry. Back to the regular programming. Those flexible mounts look quite handy, but it seems appearance can be deceiving. I'm glad I didn't succumb to temptation and buy one.

Joined: Dec 2013
Posts: 5,809
Likes: 141
DOPE
Online Confused
DOPE
Joined: Dec 2013
Posts: 5,809
Likes: 141
Originally Posted by Magnetoman
Having been in a car within a few miles of the salt flats when I was a kid, and having driven through Salt Lake City just a few years ago, I write with great authority about tuning bikes at Bonneville...


lol

i eagerly await the ariel's entry at the bub nationals and i am not joking


i'm old enough to remember when patriotism meant not trying to overthrow the government.
Joined: Nov 2011
Posts: 7,493
Likes: 528
Britbike forum member
OP Online Content
Britbike forum member
Joined: Nov 2011
Posts: 7,493
Likes: 528
Originally Posted by kevin
i eagerly await the ariel's entry at the bub nationals
Don't think the idea hasn't crossed my mind, to the extent a support trailer is already outfitted for the task.

I should have a reasonable chance of setting a record in Class SIG3A (Single-cylinder, Iron head, Gasoline, 3-speed, made by a company whose name begins with A).

Last edited by Magnetoman; 11/20/21 9:20 pm.
1 member likes this: kevin
Joined: May 2019
Posts: 976
Likes: 359
C
Britbike forum member
Offline
Britbike forum member
C
Joined: May 2019
Posts: 976
Likes: 359
You know me…. anything I can do to help. Spend your time & money, point out rabbit holes, increase your hazardous material inventory, test your patience, correct your math…..

This is a little off topic, but at one time I mentioned that I thought (based on a few sessions jetting Japanese carburetors) that the idle circuit didn’t just simply cease to function much beyond 1/4 throttle. My belief was that it still comes into play even at WOT. Basically, change was occurring in places where it shouldn’t according to all those diagrams that are still used today.
IIRC, you had determined at one point that the idle circuit still provided about 4% at WOT. Is my recollection correct and if so, how did you measure it?

Joined: Nov 2011
Posts: 7,493
Likes: 528
Britbike forum member
OP Online Content
Britbike forum member
Joined: Nov 2011
Posts: 7,493
Likes: 528
Originally Posted by Cyborg
IIRC, you had determined at one point that the idle circuit still provided about 4% at WOT. Is my recollection correct and if so, how did you measure it?
Your recollection is better than mine, so I'll have to look through old information to refresh my memory on what I might have written about that. Once I do that it could be I'll discover it's as straightforward as using the flow bench to measure the pressure drop in the pilot circuit and in the main circuit at full throttle. The ratio of those would give the ratio of fuel supplied by each.

Just writing the above has triggered a few neurons, but not quite enough to constitute an actual memory.

Joined: Nov 2011
Posts: 7,493
Likes: 528
Britbike forum member
OP Online Content
Britbike forum member
Joined: Nov 2011
Posts: 7,493
Likes: 528
Precision V-Blocks

Continuing with the sub-theme of accessories needed to make relevant precision measurements brings us next to V-blocks. Although most shafts on a motorcycle would have been machined by turning them between centers, more than a half-century of use and abuse since they were made means many of those centers no longer can be relied on to be in in good condition and/or precisely in the center.

[Linked Image]

To determine to what extent a shaft might deviate from being perfectly cylindrical, both ends can be placed on precision V-blocks and the shaft turned while a sensitive dial indicator or dial test indicator determines the total indicated runout (TIR). The next photograph shows sets of fairly small, even smaller, and even smaller still, V-blocks that are useful for measuring everything from gearbox mainshafts to carburetor needles.

[Linked Image]

Note, though, that the measured TIR depends both on the angle of the V-block, which is normally 90°, and the angular spacing of the lobes of any distortion on the shaft. Shown in the next image is the TIR of two orientations of a shaft with a three-lobe distortion (i.e. every 120°) in a 90° V-block.

[Linked Image]

In this example the TIR measured this way would be somewhat larger than the deviation of the shaft from having a circular cross-section. Also, depending on how a shaft is bent, it can be difficult to determine the precise nature and location of the problem using V-blocks. However, measurements of this type on motorcycle parts typically are used to determine if there is any measurable TIR outside of acceptable limits, not to characterize the precise value of any distortion that might be present.

Another precision measurement made possible by a V-block is shown in the next photograph, where the heavy block at the bottom of the photograph constrains the tappet from being pushed away from the indicator as the tappet is turned to measure the amount of wobble of the face.

[Linked Image]

2 members like this: Bry, Chris the camper.
Joined: Nov 2011
Posts: 7,493
Likes: 528
Britbike forum member
OP Online Content
Britbike forum member
Joined: Nov 2011
Posts: 7,493
Likes: 528
Bench Centers

Having written in my previous post that the centers in old shafts cannot be trusted, it is possible to accurately recut centers in a lathe by centering the shaft in a 4-jaw chuck and using the toolpost, not a center drill in the tailstock. For shafts that have accurate centers a bench center is an excellent tool for measuring the runout at various positions along the shaft to determine the locations of any bends or distortions in it. With accurately cut centers in a shaft, a bench center is easily capable of allowing measurement of runouts less than 0.0001" due to very slight bends.

The two indicators on the bench center in the next photograph show a BSA B34 crankshaft being measured to check progress in aligning it during assembly.

[Linked Image]

The runout on the drive-side shaft is shown being measured with a 0.001" dial indicator having 2" travel and an extended tip to clear the flywheel, a second holder is on the timing side so the dial indicator can be easily swapped between shafts, and the wobble of the timings-side flywheel is being measured with a 0.0005" dial test indicator.

Measured in the same bench center, the runout on the timing side (left) and drive side (right) of a new Phil Pearson Gold Star crankshaft is shown in the next photograph.

[Linked Image]

Runout in a straight shaft, such as a gearbox main shaft, can only be due to one or more bends in it but, in general, runout of a built-up crankshaft could be due to a bent shaft or to straight shafts that are slightly misaligned. Bench centers are very useful for locating the source(s) of runout, which helps the mechanic reduce it to an acceptable level.

1 member likes this: Bry
Joined: Nov 2011
Posts: 7,493
Likes: 528
Britbike forum member
OP Online Content
Britbike forum member
Joined: Nov 2011
Posts: 7,493
Likes: 528
Since it will take time for people to discover it hiding at the bottom of the main web page, I'll point out for anyone interested that I've started a thread documenting the complete restoration of a Vincent Black Shadow.

Surface Roughness Gauge

This thread is about measurements at length scales less than 0.001", so surface roughness gauges certainly belong since they are capable of resolving arithmetic-average roughness Ra as small as 1 µinch. The next photograph shows a surface roughness gauge mounted in an accessory variable-height holder, along with its calibration standard (124±4 µinches Ra).

[Linked Image]

This Brown & Sharpe Pocket Surf III surface roughness tester provides all the information on roughness that a motorcycle rebuilder requires. Comparable ones are made by Mitutoyo, Mahr, SPI, and others, and they are made for use by machinists on the shop floor so aren't instruments limited only to metrology labs.

Although some reamer manufacturers claim better results, more commonly you can expect the ID of a reamed valve guide to have a surface roughness of 50–100 µinches, whereas Kibblewhite recommends cast iron guides have a roughness of no worse than 32 µinches. Another example is the Parker O-ring manual recommends 10–20 µinches for shafts with dynamic seals. Without being able to measure the roughness of bushings, shafts, or ground journals you are operating blind.

An electronic surface roughness tester works by dragging a stylus with a very fine tip a fixed distance across the surface of interest, and from the up-and-down movement of that stylus the internal processor calculates various roughness parameters. These units can measure the industry-standard arithmetical average roughness Ra (which isn't the same as the root-mean-square RMS value, but in most cases it isn't too different) of a surface over the range 1–250 µinches (0.03–6.35 µm). For reference, the race of a good-quality bearing isn't even as smooth as 1 µinch (0.03 µm) so there is no component in a motorcycle that requires measurement with greater sensitivity than provided by a unit of this type.

Without going too deeply into the technicalities, the top plot in the next figure shows how a typical surface might look on the microscopic scale.

[Linked Image]

The blue curve shows the short-range roughness and the red curve the long-range waviness. What counts as "short" and "long" depends on the application, but the machining industry standard uses variations of length scale shorter than 0.030" / 0.8 mm, termed the cutoff λc, for determining the roughness of a machined item. However, although 0.030" / 0.8 mm is the common "standard" for λc, some instruments allow setting different cutoffs for specific applications. Variations on a distance scale longer than the selected λc are easily filtered out electronically, resulting in something like the bottom graph in the above figure to be evaluated by the roughness tester.

The straight blue line in the top plot in the next figure shows the arithmetic mean value of the roughness calculated from the peaks and valleys, and the distance between the black lines is the arithmetic-average deviation from that mean, designated as Ra. Although Kibblewhite, for one, lists an RMS spec for their valve guides, as mentioned earlier the root-mean-square roughness Rrms for most surfaces won't be much different than Ra so they can be used interchangeably for most practical motorcycle purposes.

[Linked Image]

The actual shape of the roughness profile depends on how a surface was prepared (milled, ground, lapped, etc.), resulting in various other roughness parameters of that are of more specialized interest (e.g. the maximum peak-to-valley height). However, for nearly all motorcycle purposes, we only care about Ra. An exception, where additional parameters would better characterize the roughness profile, would be a cylinder that had been "plateau honed" to achieve flat areas for the rings to slide against, with deep valleys to hold lubricant. Such a roughness profile is shown at the bottom of the next figure, which can be seen to look quite different than the one above it.

[Linked Image]

Although for a plateau-honed surface the single parameter Ra would be insufficient to fully characterize its flat-topped profile, it is a specialized case. As already said, with rare exception, Ra is the only parameter that is typically given, or needed, when dealing with motorcycles. For example, Sunnen lists the Ra values that can be achieved with their stones of various grits, Parker specifies the maximum Ra value a shaft should have to achieve a good O-ring seal, and Kibblewhite gives the RMS value (≈Ra) that the ID of valve guides should have.

The next photograph shows the probe of a surface roughness instrument being used to measure the stem on a used valve to determine whether it is still within specification.

[Linked Image]

At the risk of going slightly deeper than necessary into this, the way these testers do their calculations is to make a scan of total length 1, 3, or 5 times the cutoff length λc, calculate the roughness parameters within each of those basic intervals, and then average the results from the 1, 3, or 5 separate intervals to determine the final Ra.

Less accurate than an electronic tester, but still useful, is a surface roughness standard comparator plate.

[Linked Image]

The plate in this photograph has 30 specimens with different levels of roughness for common types of surfaces, e.g. a somewhat "random" roughness pattern of a lapped surface and the regularly-spaced grooves of a turned surface. In principle, this plate allows comparison with actual surfaces as fine a Ra=2 µinches. However, my eye and finger are hard pressed to discern with certainty differences in smoothness of some of these specimens below ~63 µinches. Still, a plate like this can be useful in the absence of an electronic roughness gauge.

1 member likes this: Bry
Joined: Nov 2021
Posts: 69
Likes: 3
C
Britbike forum member
Offline
Britbike forum member
C
Joined: Nov 2021
Posts: 69
Likes: 3
I don't think these have been mentioned but surface plates and tables are indispensible IMO. Taken from a Triumph workshop manual a surface table being used as a datum to check frame alignment to a very high degree of accuracy.

https://www.manualslib.com/manual/832400/Triumph-Bonneville-T120.html?page=139#manual

Joined: Nov 2011
Posts: 7,493
Likes: 528
Britbike forum member
OP Online Content
Britbike forum member
Joined: Nov 2011
Posts: 7,493
Likes: 528
Originally Posted by Chris the camper.
I don't think these have been mentioned but ...
Patience. I have a lot more material to post before I'm done.

Originally Posted by Chris the camper.
a surface table being used as a datum to check frame alignment to a very high degree of accuracy.
Nothing about that setup indicates accuracy even close to the sub-0.001" level, which is the subject of this thread.

Joined: May 2006
Posts: 371
Likes: 45
Bry Offline
Life Member
Offline
Life Member
Joined: May 2006
Posts: 371
Likes: 45
Great thread! and the post on surface roughness measurement is the best, concise explanation I have read on what can be a challenging (and usually ignored) subject for the hobby machinist.


1937 Rudge Ulster (project)
1946 Velocette MSS
1955 Triton
1959 Velocette Venom
1966 Triumph T120
1 member likes this: Magnetoman
Joined: Nov 2021
Posts: 69
Likes: 3
C
Britbike forum member
Offline
Britbike forum member
C
Joined: Nov 2021
Posts: 69
Likes: 3
Originally Posted by Magnetoman
Originally Posted by Chris the camper.
I don't think these have been mentioned but ...
Patience. I have a lot more material to post before I'm done.

Originally Posted by Chris the camper.
a surface table being used as a datum to check frame alignment to a very high degree of accuracy.
Nothing about that setup indicates accuracy even close to the sub-0.001" level, which is the subject of this thread.

Well really it's all about what is required to achieve predictability and reliability in a components/assemblies performance and is actually achievable in the typical non-temperature controlled garage. On a welded assembly of hand bent tubing being able to measure and correct deviation over a dimension of about 40" to +/- 0.001" is way better than what the factories achieved originally and enables an owner to correct what damage has occurred to a motorcycle and it's components over years of use and abuse. Is that not what this thread is about?

Last edited by Chris the camper.; 11/24/21 7:47 am.
Joined: Mar 2019
Posts: 915
Likes: 207
Member
Offline
Member
Joined: Mar 2019
Posts: 915
Likes: 207
Quote
Is that not what this thread is about?
It's actually an arms bazaar in the Tooling Wars®.


1970 T120R - 'Anton'
1970 Commando - 'Bruno'
1967 T120R - 'Caesar'
1968 Lightning - 'Dora'
2 members like this: NYBSAGUY, Magnetoman
Joined: Nov 2011
Posts: 7,493
Likes: 528
Britbike forum member
OP Online Content
Britbike forum member
Joined: Nov 2011
Posts: 7,493
Likes: 528
Originally Posted by Chris the camper.
Is that not what this thread is about?
Doing the best that is possible under a given set of circumstances always is an admirable goal. However, although this thread falls within that overall goal, it specifically deals with achieving 0.0001" absolute accuracy in the garage. While measuring to 0.001" probably sounds reasonable to most readers, measuring to 0.0001" probably sounds unachievable. So, showing that sub-0.001" on motorcycle components is achievable is what the thread is about.

Joined: May 2019
Posts: 976
Likes: 359
C
Britbike forum member
Offline
Britbike forum member
C
Joined: May 2019
Posts: 976
Likes: 359
Originally Posted by Hugh Jörgen
Quote
Is that not what this thread is about?
It's actually an arms bazaar in the Tooling Wars®.

Think you are understating it a bit…….. more like Moscow 9 May 2015.

Joined: May 2019
Posts: 976
Likes: 359
C
Britbike forum member
Offline
Britbike forum member
C
Joined: May 2019
Posts: 976
Likes: 359
I’ve thought about a surface roughness gauge, but hard to amortize the cost at this point. I’m a late bloomer and you’re a few decades ahead of me. Definitely nice to have when dealing with things like damper struts, but would like to be able to measure a plateau finish… although, given the type/quality of work it would only serve to satisfy my curiosity. Maybe something in the fax film world? Not a subject I know anything about and it probably wouldn’t be any less expensive because my microscope could have been made by Zacharias Janssen.

Although with the 3 little sponges, I could easily rationalize a decent microscope and just use a cotton ball soaked in honing fluid to check the surface finish.

Last edited by Cyborg; 11/24/21 5:15 pm.
Joined: Nov 2011
Posts: 7,493
Likes: 528
Britbike forum member
OP Online Content
Britbike forum member
Joined: Nov 2011
Posts: 7,493
Likes: 528
Originally Posted by Bry
the post on surface roughness measurement is the best, concise explanation I have read on what can be a challenging (and usually ignored) subject for the hobby machinist.
Thank you very much for the compliment. It turns out I spent a fair number of years of my career concerned with measuring sub-nm surface roughness for research on fabricating thin film structures to reflect soft x-rays. The following figure from one of my talks shows how an x-ray technique we developed was able to determine the top surface of a silver thin film had a roughness of a 2.0 nm and the substrate 0.3 nm.

[Linked Image]

Anyway, I've had a lot of practice explaining roughness measurements, and the different effects of roughness and waviness on various physical properties.

Originally Posted by Cyborg
but would like to be able to measure a plateau finish
With surfaces, smoother isn't always better, and plateau honing is an example of that. In the case of a surface against which a carbon brush rides the surface has to be smooth enough that the brush doesn't wear too rapidly, but rough enough that some abrasion does occur for good electrical contact. Industry practice calls for such surfaces to have roughness 60–80 µinches (1.5–2 µm), but if that roughness isn't random, but in the form of grooves, it can matter if the grooves are parallel or perpendicular to the direction of movement.

With the above in mind, I brought forward the description of a specialized microscope that I hadn't planned to discuss for a while.

Light Section Microscope

My microscopes are all next to each other on a bench, which makes it difficult to photograph the ones that aren't at the ends. So, the following photograph of my Zeiss light section microscope will have to do.

[Linked Image]

As indicated by the red arrows, light is reflected from the surface of the specimen at 45° into the eyepiece. The internal optics of this microscope creates a thin horizontal line of light so, if the surface were perfectly flat, what would be seen in the eyepiece would be just a thin horizontal light. However, if the surface is rough, the reflected line will take on the shape of that roughness. I can't find an image I took of a magneto's earthing ring so the following figure comes from the microscope's manual.

[Linked Image]

As I've noted on this figure, the field of view is 0.03" wide, and peaks and valleys of the roughness as small as 20 µinches can be resolved. A reticle moved by a calibrated knob allows the height at any horizontal location to be measured.

The information from this specialized microscope is complementary to that from an electronic surface roughness gauge. A gauge calculates the average roughness (which, in principle, could be painstakingly calculated manually from the above photograph) to compare with a desired specification, but the microscope allows the actual profile of that roughness to be seen. The electronic gauge would tell you the average roughness of a part made on the lathe was, say, 60 µinches, but if individual grooves were seen in the image from the light section microscope you would know you could achieve a smoother surface by simply substituting a different cutter with a larger radius tip and/or reducing the feed rate.

1 member likes this: Bry
Joined: Mar 2019
Posts: 915
Likes: 207
Member
Offline
Member
Joined: Mar 2019
Posts: 915
Likes: 207
Isn't that just like a wop?
Brings a microscope to a micrometer fight.


1970 T120R - 'Anton'
1970 Commando - 'Bruno'
1967 T120R - 'Caesar'
1968 Lightning - 'Dora'
Joined: Nov 2011
Posts: 7,493
Likes: 528
Britbike forum member
OP Online Content
Britbike forum member
Joined: Nov 2011
Posts: 7,493
Likes: 528
Again, Cyborg gets the blame for another instrument having been moved forward in the discussion. It's his fault for mentioning plateau honing... OK, true that I mentioned it first, but it's his fault for mentioning it again.

First, a discussion of Sunnen hones, followed by plateau honing.

Sunnen Hone

While a hone may not immediately come to mind when thinking about precision tools, it is capable of sizing a hole to approximately 0.0001". This is implied by the dial on the MBB-1660 Sunnen hone, shown being used on a valve guide being fabricated for a BSA.

[Linked Image]

The dial on the Sunnen displays the change in radius, not diameter. However, very roughly speaking, approximately half the movement of the dial is typically due to material that has been removed from the part and the other half due to loss of material from the honing stone. Knowing this, and having some small degree of experience with the hone, allows control over diameters to ~0.0001"

The next photograph shows a bush at the point when another 0.0003" remained to be removed from the diameter of a bush, a precision that the Sunnen hone is able to achieve.

[Linked Image]

In the past, Sunnen made hones of this type for the automotive as well as the industrial market. However, the former have an upper speed of only 640 rpm that isn't optimum for holes smaller than 1", while the latter with a 2500 rpm upper speed is much more versatile for motorcycle work. Although both types made in the past 60+ years look much the same, the model numbers of the lower-speed automotive hones have LBB prefixes and almost always are painted red, while industrial hones are MBB and have been green, grey, or ivory at different times.

Unfortunately, with the exception of cylinders, each mandrel only covers a limited range, so a large number of them would be needed to cover the entire range with no gaps. For motorcycle use mandrels are available for honing everything from a 0.105" needle jet to a 3.39" Matchless G80 cylinder, with the AN-600 mandrel covering the range from a BSA C15 on up.

[Linked Image]

"Standard" mandrels with sizes up to 1.25"-diameter hold one stone of length that depends on the size of the mandrel, with the stones 1"-long for a 0.106" mandrel and increasing to 4½"-long for the largest of these one-stone mandrels. However, starting at a diameter of 0.619", mandrels holding multiple stones are available. For example, a 3-stone ⅞" mandrel covers a length of 7½". Since the distance from the outside of the main bearing to the outside of the ⅞" timing-side bush on my Ariel is 5.6" this means a 3-stone mandrel can be used to line-hone the bush for accurate alignment with the bearing (of course, using a faux main bearing on the drive side for the honing). Such a multi-stone mandrel would be useful for the same purpose on a BSA A10 engine.

Sunnen also makes external hones in three sizes covering from 0.120" to 4.5" for smoothing the surfaces of shafts.

[Linked Image]

While the MBB-1660 can be used to hone guides when they are installed in a head, the angle and weight can make doing so difficult. Instead, often a portable Sunnen P-190 Honall is a better choice.

[Linked Image]

Although the adjusting knob on this hone is uncalibrated, it isn't difficult to achieve 0.0001" accuracy on the ID of holes with it.

OK, back to plateau honing. By modern standards our cylinders are honed with very coarse grit (i.e. ~180) to seat the rings, so it's not clear plateau honing would be a good idea. Still, if that's what someone wanted, the top figure in the following shows what the surface might be like after honing with 180 grit, and the red line shows where we might like the top of that surface to be in order to be smooth for the rings, but with valleys to hold lubricant.

[Linked Image]

According to Sunnen their 180 grit stones would leave an Ra of ~30 µinches. The problem is simply that of removing everything above the red line while leaving the valleys below the red line as they are, as shown in the bottom figure of the above.

If we honed the cylinder a second time "long enough" with 500 grit we would end up with 5 µinches across the entire surface which, for present purposes, I'll call perfectly flat. Conveniently, Sunnen provides a formula that tells us how long "long enough" Is. To go from one finish to another requires

required stock removal = [existing finish – desired finish] / 100,000

So, to go from Ra = 30 µinches everywhere to 5 µinches would require removing 0.00025". That means if we remove only, say, 0.0001", an amount which can be controlled by the Sunnen and measured by bore micrometers, it would leave a flat surface with valleys to hold the lubricant, i.e. a plateau finish.

The above describes one approach to achieving a plateau finish if someone didn't have an electronic roughness gauge, but with a gauge a more controlled approach is possible. In addition to measuring the average roughness Ra, a flip of the switch allows display of the maximum peak-to-valley distance Rmax in a scan of length ln, or the average Rz of the five individual Rz values in the five λc lengths that make up that scan, as shown at the top of the next figure.

[Linked Image]

So, the procedure would be to measure Rz after honing with 180 grit stones, then install 500-grit stones and hone until Rz had been reduced by a factor of two (or whatever amount you decide), The 500-grit stones only would abrade away the peaks while leaving the valleys untouched, as shown at the bottom of the above figure.

Last edited by Magnetoman; 11/25/21 5:02 pm. Reason: added text about multi-stone mandrels
Joined: Oct 2012
Posts: 6,042
Likes: 182
knuckle head
Offline
knuckle head
Joined: Oct 2012
Posts: 6,042
Likes: 182
I have the shop use plateau honing on Triumph cylinders with a finer stones than most use. No problems breaking in cast iron rings. My land race Triumphs use gas nitrited nodular iron top rings with 320 finish and again, no problems..
When the race bike was the first to set a modified production 650 gas pushrod record over 130 mph I had freshened up the engine ...The bores and pistons were scratched from no air filters. Removed the rings, rubbed the pistons with a green scrubbie pad... Then with the cylinders held by my boots on an old cookie pan placed on the shop floor, hone the cylinders with a brush hone using brake fluid as a lube.Then ran the drill slow in reverse for a few revolutions for a plateau finish....Perhaps at 7400 rpm with the crank flexing and fretting parts, precision machining is no longer a factor... grin


79 T140D, 89 Honda 650NT ,61 A10 .On a bike you can out run the demons
"I don't know what the world may need
But a V8 engine is a good start for me
Think I'll drive to find a place, to be surly"
2 members like this: Magnetoman, NickL
Page 4 of 6 1 2 3 4 5 6

Moderated by  Richrd 

Link Copied to Clipboard
Job CycleBritish Cycle SupplyMorries PlaceKlempf British PartsPodtronicVintage MagazineBSA Unit SinglesBritBike SponsorBritish Tools & FastenersBritBike SponsorBritBike Sponsor






© 1996-2021 britbike.com
Powered by UBB.threads™ PHP Forum Software 7.7.5