View Full Version : Timber grading

Alan Case
June 22nd, 2004, 06:33 PM
Hi, I know there are a lot of people on this site with experience in the timber industry, so I wish to draw on your knowledge. I am presently working in a timber mill where the timber is visually graded by the operators who place a chalk mark (different symbol for different grades) on the timber to indicate the grade and dock required. The timber then goes through a lucydyne scanner which reads the mark and returns a combination of digital inputs to the PLC. The operators are very slack with the quality of their marks (due to the timber moving past) which means there are a lot of misgrades. I am wondering how other plants do this process, ie has anyone tried a mark in different areas of the board to indicate grade.
That is put 4 divisions on the board with laser beams (3 laser beams 4" apart) and a chalk mark anywhere in that area of any pattern to indicate the grade.
I await any ideas or suggestions. Regards Alan Case

Eric Nelson
June 22nd, 2004, 07:45 PM
I have zero experience in the timber industry beyond the lumber aisle of Home Depot, but anyway... :D

How about marking the timber with a UV marking pen (http://www.sick.com/us/products/new/csm1/en.html), then reading the marks with a UV sensor (http://www.sickusa.com/live/master/datasheet.asp?PN=1023501&FAM=Lumienscence)?... :huh:

The links I provided are just examples I found in a quick Google search.



June 22nd, 2004, 07:51 PM
Could they put stickers with barcodes instead of chalk marks?
Maybe Color code the Stickers???

BTW I have as much experience in the Lumber industry as Eric.

Sometimes ideas out of left field help lead us to a better solution.

June 22nd, 2004, 08:31 PM
How about a system with a few push buttons. The operator would press a button corresponding to the grade chosen. Each button would activate a bar code printer which would print the appropriate code on the piece. Then the bar code reader could read the code.

June 22nd, 2004, 09:41 PM
I did something similar a few years ago with tires.

Several factors were invloved.

The weight of the tire would automatically send it down a chute.

It would be balanced and maybe sent down a chute.

Physical inspecion could divert it.

When it was all done, there was a rack of perfect tires, and depending on the number of marks imprinted along the way by "dot placers" and the location of the mark placement, determined were the rest of the tires went, and which store got what was supposed to be identical "good" tires. Most marks were read with a photoeye looking for a "light" indication.

June 22nd, 2004, 09:44 PM
You could possibly use paint. I worked on a plant where they finger joint timber to make long lengths. At each end of the timber a blue mark was sprayed which a Sick colour sensor picked up on. When the sensor saw the mark the timber was clamped and pressed together.

Maybe you could have a grading station where the operator could grade the timber and select the colour to spray for the grade via push buttons. The spray duration could be controlled by a timer.

Peter Nachtwey
June 22nd, 2004, 11:44 PM
However, I don't agree with the basic idea that marking lumber is any better than pushing buttons. Writing is much harder to do that pushing buttons. I don't know of any scanning system that can grade as well as a person and that is the problem. Even back in the 80s when I was doing lumber sorters we had a high a low grading line so that there were two people grading lumber instead of one. I think having two separate lines was a little costly but it didn't take much space as one as above the other. Another idea would be to have multiple trimmer/grader people grading every 3 or 4th board. A colored light would be required to indicate which board a particular grader should grade. This would reduce the load by the number of graders. However the labor costs go up.

I tried grading lumber for just a few minutes in the 80s. You guys can believe how mind numbing it is grading a board every second or faster. It is not easy because it requires total attention. Try it and see.

There are no simple answers. The ultimate answer would be to get one machine that can do the grading without human assistance but that may be a few years out yet. I would keep in touch with the leader in scanning technololgy in the industry. The Lucidyne guys are pretty sharp and are always push technology.

BTW. I worked with most of the senior guys now at Lucidyne in the 80s before Lucidyne was started. Its a small world. I wonder if any of our hydraulic motion controller are in your mill.

Alan Case
June 23rd, 2004, 04:47 AM
Thanks to all. They use 5 graders at present and each grader marks approx every fifth board. As there are 90 boards going past every minute then there is no time to place stickers etc. I might do a bit of research into the Lucydyne products as I think the have a lower end model. Colour not recognised.

Peter, what brand are your controllers? I will check what there is here. The whole mill will be demolished in 4 months time as the new mill comes on line so if they are on site I will have to be quick. After Christmas when they reach full production they will be the largest in the southern hemisphere and fifth largest in the world (so they tell me)

What system of marking did you find worked the best. They are trying to put a 6 on the timber here as a grade mark and it is not too successful.

Regards Alan Case

June 23rd, 2004, 08:17 AM
This may be a little bizarre but bear with me, it's past my bedtime.

IF the boards could be fixed in place on the conveyor: an overhead camera scans the marking area. Each inspector 'marks' the bad sections with a laser or other light source which the camera can detect. When the camera 'sees' an illuminated defect area, it can synchronize that spot with the conveyor movement and something downstream can apply a definite, consistent paint mark for the cutter. Think along the lines of 'painting' a target for a laser guided bomb.

June 23rd, 2004, 06:45 PM
I believe that grading occurs downstream of trim saws and only determines where the lumber should be stacked when taken off the "green chain". My experience pre-dates automatic green chains. When I was a "puller" on the green chain, we read the mark ( a crayon 'B', 'E', 'S', 'C', etc.) and stacked in separate piles according to mark and length. The graders have to inspect both sides of a board, but occasionally miss. More than once we would pull a seemingly good board only to find the underside full of teredoes - yuk! Then you'd have to handle it again to throw it back on the chain so it could go to the reman pile.
Do you get teredoes in Oz?

June 24th, 2004, 01:03 AM

We use a newnes grademark reader (they brought Luncidyne out)
apparently they have improved the grademark readers quite a lot and they seem to be very flexible with the reading of the grade marks.

Im quite impressed with ours.

Are you sure its not a light level problem with the uv tubes or
the back bar reference strip is worn/dirty?

this can have a big effect on the grade mark interpretation.

Our old mill had a combination of pe cells and push buttosn to determine the visual grade.

the stress grade was colour blazed on the board and detected on the lug chain just after the trim saws with a colour sensor whic in turn operated our inkjet printers.

hope that helps

Alan Case
June 24th, 2004, 08:14 AM
Hi, the grading is done before the trim saws as part of the grading mark is what dock is required. The grading is the only time that the boards are touched as everything else is fully automatic (except for misgraded boards having to be pulled out at the stacker)
What is a teredo? Regards

John Rhea
June 30th, 2005, 12:54 AM
Hi, I work for Lucidyne Technologies, Inc. and have for about 15 years.

The technology today does include systems that can grade lumber for both geometric and biological defects. The cost of these systems can be quite expensive.

In response to one reply, Lucidyne has not been purchased by Newnes. The purchased the Peter Appleton system.

The manual marking with any system is going to only as good as the people doing the lumber marking.So, it is very important to use markes they are easy to make and will not be confused with other marks that can lead to misgraded lumber. Take time to review these marks and impress upon the graders how important their job is. I have seen many mills that run with less the 1/2% rejects.

June 30th, 2005, 02:07 AM
Hi Alan,

Actually I built an entire optimising cutoff saw using and SLC5/03 and IMC110 about 10 years ago, using just such a chalk and UV sensor arrangement. I agree with you, it is totally dependent on the quality of the grading operators marking to make it work.

June 30th, 2005, 11:15 AM
The graders are very important and yes it is a numbing mindless job. You become a robot thinking only about a long list of critera for each board. My wrist hurt after 2 hours. ( you have to turn each board). I use to watch when rejects started getting high. Many times the problem was the flash tubes, not the grader. Even if they are working, they might not be as bright and may need replacing. Luci is a high maintenance tool. Also ambient light getting inside the camera area can wash the image. If you watch the monitor you may see the odd faded image. Working right, everything in good shape, graders keeping alert, you should be over 97%. Lucidyne is one of the better grade readers. I am now off work and going to enjoy a long weekend.... :) :) Have a great weekend!

July 2nd, 2005, 08:42 AM
My past experence with "lucy" and the human set of lumber graders I've worked with has lead me to believe that the "6" mark is not a very reliable mark. The solutions I've seen have ranged from not using "6" as a mark to using it on a very limited selection. The limited selection could be a grade that had only one length or width that was unique from the rest of the product being run. Then the sorting system (presuming there is width and length sorting) can help to identify that lousy "6" mark correctly. The if you have a "6" on a board that does not meet the correct sorting criteria it becomes a reject (do you have a reject bin?). This way the boards don't have to be pulled at the stacker.
But perhaps the most effective tool I've seen for getting good penmanship on the boards is to assign a color to each grader (making them responsible for their own marks). Then if one color is always needing to be pulled at the stacker, management can apply the the correct feedback .

July 2nd, 2005, 10:42 AM

My experience is limited to making modifications on an existing lumber sorter/grading system. This one was built by CSMI (the old Consolidated Sawmills, Inc). It has two grading stations mounted above the incoming lumber chain. The chain splits, with one half going to the upper grading chain, and one half to the lower. The graders each sit in a booth where they have bright lights aimed at the boards. The boards move down an inlined chain section that puts the board directly in front of the graders, and the grader has a button that will flip the board over so that both sides can be seen. Then the grader has about 2 seconds to make his decision and hit a grade button on a console near his right arm. When the button is hit, the PLC5 assigns this grade to that lug on the chain. The chain has an encoder, so that the PLC, using Bit Shift Registers, keeps up with where the board is as it goes through the trim saws, and finally to the assigned bay in the storage area.

In this system, it is not necessary for the graders to make marks on the boards. The PLC keeps up with the grade and puts the board in the right storage bay according to wood type, length, thickness, and grade. There is a paint marker that is operated by the PLC. I think that this is used to mark a board as to whether it passed the minimum thickness test. The last bay under the main chain is the Reject bay, for boards that cannot be used for any purporse.

Terry Woods
July 2nd, 2005, 11:10 AM
Manually grading boards at one location is tough... really tough. This becomes tougher as line speed increases. This becomes even tougher still as the number of grades increases.

Let's say, just for the heck of it, that one particular line might handle several grades. The grader has to make a quick one-of-several evaluation and then provide the appropriate indication (mark). A "reader" then has to decipher the indication to make the correct decision regarding destination.

The whole exercise is mentally and physically extensive. Not that there is any particularly difficult mental activity or physical activity, it's just that it's rapid and repetitive. It's got all the makings of a Zombie-Generator.

The "reader" requires a particular set of conditions to provide accurate interpretation of the markings.

Now, just for the sake of discussion...

The mental and physical stresses of grading can be reduced significantly if any given grader only has to decide between "GOOD" and "NO-GOOD". At any stage in the grading game, the particular grader is looking for only the best. The grader selects the boards that he wants by marking them for selection. The board is either marked or not marked. A relatively simple "reader" then looks for the presence of the mark. If a mark is seen then the board is diverted. All boards that are not diverted are passed on to the next grading station.

At the next station, it is again, simply a case of GOOD or NO-GOOD. Just as the previous grader did, this grader is looking for only the best. The board is marked, or not marked, accordingly.

At each station, the "cream is being skimmed from the top". This is repeated for as many grades as there are until all that's left is Hog-Fuel (these boards are ground-up and sent to the steam plant as boiler fuel).

At any station, evaluation is based simply on GOOD or NO-GOOD.
Mark detection is based simply on the presence or non-presence of a very simple mark... perhaps a single line.
Marks do not have to be "color-based".
Mis-grading is virtually eliminated. (Worst case, a better board is passed to the next lower grade.)
Graders do not have to physically pull boards.
(Boards that would be pulled are simply passed to the end of the line).
This can be developed with a single grading line.
Grader positions determined by experience... Rookies at the end.
The zombie-factor is greatly reduced.

Grader performance is determined by the content of his bin. As long as a particular bin has no lower-grade boards, then the particular grader is doing well. If a particular bin has boards of a higher quality than it should, then the previous grader is not doing so good.

This scheme evaluates boards more by "what they are" rather than "what they are not".

Multiple Graders... $$$$.

It becomes a question of cost. Certainly there is the cost of multiple graders. But then that has to be compared to the loses caused by initial mis-grading and the cost of more sophisticated Mark Detection equipment and the loses caused by mis-interpreted grade marks... and, of course, the cost of down-time in the event of any Mark Detection equipment failure.

Peter Nachtwey
July 2nd, 2005, 11:47 AM
I know that board turners have existed for years.

I think it would be best if there was a camera that can view one side of the board, flip the board, and then another down stream camera views the other side and then the computer grades the board and ONE grader has the option to override the the computer. This information can then be passed on to the trimmer.

From what I am reading, it doesn't sound like the computers can grade the boards yet. They can only read marks.

July 2nd, 2005, 12:49 PM
Ok my hairbrained idea:

why not color code the lumber? The inspector holds a remote with 5 buttons, each button will cause a paint gun to spray a different color, indicating grade, on the lumber as it zooms by.

BTW, this sounds like a terrible job. I hope they do it in short shifts..

Terry Woods
July 2nd, 2005, 08:59 PM
Don't forget...

I said... "Manually grading boards at one location is tough... really tough."

And... the original question was posed in terms of Manual Grading.

So many people are so willing to throw your money at their particular solution.

Got a problem? Spend Money! Forget about evaluating the true nature of the problem and possible solutions using your current configuration!


If that doesn't work.... well... hmmm... Ah! Yes! Of course! That's it!... SPEND EVEN MORE MONEY!!! That's really the CAPITALIST WAY! Spend more money after bad money.

Sooner or later you'll either spend your way to a solution... or, you'll spend your company out of existence.

Either way, the "solution provider" makes out.

Something wrong with this picture?

THINK YOUR WAY OUT OF PROBLEMS! Don't try to spend your way out of problems unless you really have no other choice! Think First!

It's called... Common Sense!

July 2nd, 2005, 09:27 PM
This is not a new problem. Asr you trying to increase production? Try colored markers for each grader and track their lumber grades. Shouldn't the shift grades be similar? Scanners are not good for internal defects, but are good for lumber profile problems. The Grader can see pitch pocket, knots... Would a scanner help in your recovery? Calculate the return on investment.

July 3rd, 2005, 07:34 PM
Basically the topíc guide for general exposition of classification wood cut and peeled. He has some condition to pursue?

However, I don't agree with the basic idea that marking lumber is any better than pushing buttons. Writing is much harder to do that pushing buttons.

Because, that it becomes complex?

Detention imperfeitions in wood [surface]
Condition: Salient imperfections or coarse.
Sense:Mechanical parallel -Device with free wheel (small diameter)in bottom part with vertical axis + sense of light or encoder.
Operation: The turn of the wheel makes to produce pulses, when it moves on an imperfection.

Somebody knows on a cientific study on the retention or disturbance of radiation that of the imperfections of the wood?
I read to folder of manufacturer, that deduced to use something it sort, whom had the expression "latent detention", a machine to cut imperfections