Maxcess International -Fife, MAGPOWR, and Tidland present Converting Blog

« Home » Tensioning, Web Mechanics

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 09:00 AM

Pneumatic systems are popular on low cost converting equipment. Using a brake or clutch and an air pressure gauge avoids sophisticated electronics, wiring, and control systems.

If you set web tension from an air pressure gauge, there are two likely possibilities of what you are doing. If you have a dancer roller, you may be controlling the air pressure to one or two air cylinders that apply an external force to the dancer roller. If you don't have a dancer roller, the air pressure is likely going to a pneumatic brake or clutch, setting the slipping torque transmitted to the unwinding roll, winding roll, or a torque-controlled roller.

In either case, I highly advise spending some time with a force gauge (like a fish weighing scale) and measure the relationship between the air pressure settings and web tension.

[This entry is part of a growing 'book' that I am writing and posting at www.webhandling.com. Convertingblog.com is getting an exclusive first look at this restricted material.]

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 01:18 PM

Definition: Elasticity is the property of returning to an initial form after deformation.

Rubber is the classic example of an elastic material. You push on it, it deforms; you let go, it bounces back. An elastic material responds to load almost immediately (the load travels through the material at the speed of sound). The amount of deformation is proportional to load and independent of time.

Definition: Viscous is the property of having a relatively high resistance to flow.

Molasses is the classic viscous material. When a force is applied to a viscous material, it will flow. The longer the load is on the viscous material, the more it will flow. When the force if removed, they stop flowing, but won't recover.

Definition: Viscoelasticity is the property of having both viscous and elastic properties.

When a viscoelastic (V-E) material is loaded, it will respond with a mixture of viscous and elastic behavior. Upon loading, a V-E material will immediately stretch (elastic behavior) and begin to flow (viscous behavior). When the load is removed from a V-E material, it will recover, some immediately (elastic behavior) and recover more over time (viscous behavior). Vinyl electrical tape is a classic and easily observed V-E material.

Do this test yourself. Pull out a 2-3 foot length of electrical tape. Hang a 1-2 lb weight on it. Note the initial elongation and that the tape will continue to elongate. Take the weight off. Note the initial recover and ongoing recovery.

Congratulations, you've just completed your first creep test.

[This entry is part of a growing 'book' that I am writing and posting at www.webhandling.com. Convertingblog.com is getting an exclusive first look at this restricted material.]

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 09:00 AM

Q: Which is the best direction for a dancer roller's motion, horizontal or vertical? Also, what is better, a linear or pivot dancer motion?

A: I don’t have a strong preference on horizontal vs. vertical, but I am strongly in favor of pivoting dancers over linear dancers.

I prefer a vertical dancer arm with horizontal motion so that gravity is perpendicular to the tensioning direction, but make sure the dancer arm is hanging (as opposed to standing) with the the pivot point above the roller. Hanging dancers are 'sloped' to run stable; gravity tends to pull the roller back to its null position. Standing dancers, with the pivot underneath the roller are destabilizing with gravity trying to pull the roller away from neutral position.

My preference for pivoting over linear is based on observation of friction and sticking almost always higher in linear systems. If well-engineered, there is no reason a linear system can work great, but too often a linear dancer has more trouble holding its alignment, creating tension variations and wrinkles. A pivoting dancer can also be poorly engineered, but you have a better change to get low friction in a rotary bearing and simple structural design can translate a well-aligned pivot shaft into a well-aligned roller.

-tjw

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 10:56 AM

How do you measure a baggy web?

A common, practical, but tough question. I suggest several options, some off-line, others on-line.

1. Measure skew from a straight line in sweeping out a web (measure asymmetry only).
2. Measure crossweb droop variations in a lightly tensioned horizontal span.
3. Qualitatively judge a sheet of film as you pull it from a roll, grading it a 1, 2,..5 as you pull a web to tautness with finger, wrist, elbow, shoulder, or full body strength.
4. Combine 2 and 3 above, measuring tension to pull horizontal span up to less than X (say 1") droop.
5. Mark crossweb lines in a wide web, then cut it into machine direction strips and measure the length differential of the strips.
6. Lay flat test, place sheet on flat plate and measure ripple or curl deviations from planar.

On-line tests (measuring the moving web):
1. Droop test with scanning distance measurement, such as laser triangulation micrometer or ultrasonic sensor.
2. Low angle laser - Visualize droop with a low-angle linear laser or combine this with a camera and vision system to quantify bagginess (I don't know of any commercially available system to do this.)
3. Multi-segment tension roller - Dover flexo makes this. Segmented tension roller.
4. Multi-segment tension bar - Same idea, but for scratch-insensitive webs, using a segmented bar instead of rolling elements. (I don't know of a commercially available design, but Paprican will talk about this at the AWEB next week.)
5. Air turn back pressure - Float the web over an air bar and measure the back pressure in discrete positions across the web width. (I haven't seen this commercially available in the US, but I expect any air turn manufacturer could put one together, such as MEGTEC. I have seen this marketed in Japan by Bellmatic Ltd.
6. Speed of sound - Measure the speed of a sound wave in the web. First demonstrated by Dr. Richard Lowery at the OSU WHRC. Limited marketing of this device by Ron Markum of Advanced Web Systems.

More. I don't think any of these have completely satisfied the need. Let me know if you want to add to the list.

tjw

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 02:06 PM

A few months back, I made post asking 'Is motor torque control EVER the right?' This was a post based on my experience with several converting controls engineers.

Pete Werner, formerly of Rockwell Automation, posted his comments on this and made a strong case of torque-based motor control should be the first choice. This is quite a contrast from what I've seen over the last 20 years.

Since then, Pete has continued to educate me on this topic. Pete made an interesting presentation at the AWEB conference last month. He makes some good points about how speed regulation and dancer rollers the popular choice, but these are devices invented to solve problem many years ago.

Why does speed-based tension control continue to dominate converting applications?

Probably several reasons. First, controls engineers have learn this and know it works. Speed is easy to understand.

Why use speed reducers (gearboxes)?

Supplier like to boast speed accuracy. Controls engineers like using speed reducers to reduce motor size and increase efficiency. But you shouldn't be thinking about energy or motor costs when your goal is good tension control. Yes, eliminating the speed reduce increase motor size and cost (both torque related), but you save some of that back by not buy a gearbox. The most important motive should be good (or great) tension control.

I always had it drilled into me that you want a motor to run using its full operating speed range. If the motor has 0.01 percent speed accuracy, this is 0.2 rpm variation over 2000 rpms. If you run the motor at 100 rpm with no speed reducer, you're asking for 0.2 percent speed variations. Why would you do that?

This is the wrong question. Motor are not inherently speed devices. They are torque generators. Torque output divided by the mechanical leverage of radius creates tension. Why care about speed accuracy at all?
These are great questions.

I like Pete's question where he says 'You don't usually consider putting a speed reducer on a braked or clutched application, why do it on a motor?'

Today's drives call for new answers. I get the feeling Pete would like to eliminate 90 percent of the speed reducer (gearboxes) and dancer roller in the converting industry.

Let me know your thoughts or experiences regarding speed or torque motor control and when dancer rollers can be eliminated.

tjw

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 10:20 AM

I just rec'd an amazing testimonial from Dr. J. Keith Good at OK-State U's Web Handling Research Center (WHRC). Can you believe motor controlled machine with a 6000:1 controllable speed range?

I created a little stir a few months back in asking whether torque control is EVER the right choice for closed-loop tension control. I didn't think it was, mostly due to the natural torque losses of speed reducers getting in the way of good torque transmission.

Pete Werner, formerly of Rockwell Automation, has been educating me on how torque control should be the first choice... and get rid of those speed reducers. Can we believe Pete? Well, I think YES. Read this testimonial from Dr. J. Keith Good, top researcher at the WHRC.
=====================================================
Tim,
Rockwell has done a superb job on our High Speed Web Line. It now has the Flex 700S phase II drives with the through axis Heidenhain encoders. This requires a special option board on the 700s to communicate with the encoder. I have no speed reduction in any form today, we are direct drive 1:1.

When our line was originally commissioned with Reliance's 1st generation VCI controllers cogging was a problem when running at low speed. In that day performance specs stated you should not run at less than 10% of full speed, thus we should not have hoped to have good speed and tension control until we reached 500 fpm. If the drive technology had not improved I was considering 2 speed gear boxes similar to Art's old solvent line to allow me to run at 50 fpm without cogging.

Today I'm running the same motors the line was commissioned with but with the new drives and encoders listed above. Tim I can run smoothly with no cogging at 1/2 fpm and if I had the guts I could go 5000 fpm. I don't because I still have some idlers that scare me at 3000 fpm. Also we commonly command zero fpm in our experiments. Tension and drives are fully on holding command tensions at zero velocity. We then transition to 10 or 50 fpm and the machine does so beautifully.
Thus I believe it is now possible to run at low speeds without cogging. I have yet to meet a zero backlash gearbox or right angle gearbox (no matter what their written specs are) so if you don't have to have them I'm all for eliminating them.

Best regards,
Keith (Dr. J. Keith Good, OSU)

=================================================
Maybe 10000:1 speed range, with a little guts? Whew.

BTW, I don't mean to make convertingblog.com a commercial promotion website, but I don't see a reason to leave out the brand name of products that are key part of a great solution story that help raise the knowledge of the entire converting community. I expect you can get similar results from non-Rockwell products and I encourage anyone to share their experiences here.

I happy to name brand name's on the good stories. I'll likely remove brand names where it is needed 'to protect the innocent.'

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 09:38 PM

A few months back, I made post asking 'Is motor torque control EVER the right?' This was a post based on my experience with several converting controls engineers.

Pete Werner, formerly of Rockwell Automation, posted his comments on this and made a strong case of torque-based motor control should be the first choice. This is quite a contrast from what I've seen over the last 20 years.

Since then, Pete has continued to educate me on this topic. Pete made an interesting presentation at the AWEB conference last month. He makes some good points about how speed regulation and dancer rollers the popular choice, but these are devices invented to solve problem many years ago.

Why does speed-based tension control continue to dominate converting applications?

Probably several reasons. First, controls engineers have learn this and know it works. Speed is easy to understand.

Why use speed reducers (gearboxes)?

Supplier like to boast speed accuracy. Controls engineers like using speed reducers to reduce motor size and increase efficiency. But you shouldn't be thinking about energy or motor costs when your goal is good tension control. Yes, eliminating the speed reduce increase motor size and cost (both torque related), but you save some of that back by not buy a gearbox. The most important motive should be good (or great) tension control.

I always had it drilled into me that you want a motor to run using its full operating speed range. If the motor has 0.01 percent speed accuracy, this is 0.2 rpm variation over 2000 rpms. If you run the motor at 100 rpm with no speed reducer, you're asking for 0.2 percent speed variations. Why would you do that?

This is the wrong question. Motor are not inherently speed devices. They are torque generators. Torque output divided by the mechanical leverage of radius creates tension. Why care about speed accuracy at all?
These are great questions.

I like Pete's question where he says 'You don't usually consider putting a speed reducer on a braked or clutched application, why do it on a motor?'

Today's drives call for new answers. I get the feeling Pete would like to eliminate 90 percent of the speed reducer (gearboxes) and dancer roller in the converting industry.

Let me know your thoughts or experiences regarding speed or torque motor control and when dancer rollers can be eliminated.

tjw

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 08:26 AM

How do you measure a baggy web?

A common, practical, but tough question. I suggest several options, some off-line, others on-line.

1. Measure skew from a straight line in sweeping out a web (measure asymmetry only).
2. Measure crossweb droop variations in a lightly tensioned horizontal span.
3. Qualitatively judge a sheet of film as you pull it from a roll, grading it a 1, 2,..5 as you pull a web to tautness with finger, wrist, elbow, shoulder, or full body strength.
4. Combine 2 and 3 above, measuring tension to pull horizontal span up to less than X (say 1") droop.
5. Mark crossweb lines in a wide web, then cut it into machine direction strips and measure the length differential of the strips.
6. Lay flat test, place sheet on flat plate and measure ripple or curl deviations from planar.

On-line tests (measuring the moving web):
1. Droop test with scanning distance measurement, such as laser triangulation micrometer or ultrasonic sensor.
2. Low angle laser - Visualize droop with a low-angle linear laser or combine this with a camera and vision system to quantify bagginess (I don't know of any commercially available system to do this.)
3. Multi-segment tension roller - Dover flexo makes this. Segmented tension roller.
4. Multi-segment tension bar - Same idea, but for scratch-insensitive webs, using a segmented bar instead of rolling elements. (I don't know of a commercially available design, but Paprican will talk about this at the AWEB next week.)
5. Air turn back pressure - Float the web over an air bar and measure the back pressure in discrete positions across the web width. (I haven't seen this commercially available in the US, but I expect any air turn manufacturer could put one together, such as MEGTEC. I have seen this marketed in Japan by Bellmatic Ltd.
6. Speed of sound - Measure the speed of a sound wave in the web. First demonstrated by Dr. Richard Lowery at the OSU WHRC. Limited marketing of this device by Ron Markum of Advanced Web Systems.

More. I don't think any of these have completely satisfied the need. Let me know if you want to add to the list.

tjw

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 07:24 AM

This is a question I'm often asked by prospective clients for my services as a web handling specialist.

Have you ever worked with thin webitronium? (Yes, this is a ficticious material.) Have you ever handled nylon film? Have you ever worked with nickel-steel foil? Can you handle my complex laminate?

My answer is "I don't care what you call it, let's talk about the material properties:
elastic modulus (in x, y, and z directions),
thickness,
width,
speed,
roll sizes (core diameter, final diameter)
surface friction or adhesion,
expansion coefficients (thermal or moisture), roughness,
break strength,
yield strain-strength (in x, y, z directions),
conductivity,
etc.

Once these are defined, it doesn't matter what name you want to give a product. Call it Fred if you want to. Once I know this mechanical data, we can apply engineering and physics to solve your problem.

I hope this doesn't sound too theoretical. I'm also quite practical. I like to apply what works. One thing you get from bringing in someone from outside your industry is a solution that may already be proven elsewhere that can solve your problem. In that regard, I'm a honeybee that cross-pollenates technology from one industry to another (of course, the non-proprietary knowledge, only).

I look forward to handling Fred and preventing his wrinkles.

tjw

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 09:38 AM

What is the best tension control plan for processes that start and stop frequently?

The biggest problem with stopping and starting is inertia and coordinating motor speeds, especially for the unwinding and winding rolls. Inertia time acceleration (or deceleration) creates a torque (at steady speed, there is no accel, so no torque). The inertial torques of rolls or rollers can easily be many time the torque of tension (times radius of the roller) or other torque losses (drag, etc.).

Any tension control plan must answer four questions:
Q1. How will I apply torque to the web to create tension (motor, brake, clutch)?
Q2. How many tension zones will I have in my process?
Q3. Which drive point will be my master or pacer drive (the motor that is in speed control, no tension loop)?
Q4. How will I trim the torque or speed of the follower sections?

Stop start may affect all of these decisions.
A1. Regenerative motors (motors that can function in power or braking mode) can be programmed to respond to predicted inertial loads, reducing tension upsets.
A2. The number of zone you need may go up with high inertial torque from a process with a large number of rollers.
A3. You need to ensure that the pacer section doesn't slip. The inertial loads of stop-start processes can create surprising tension changes within a zone and across drive points. Becare to estimate the tension swings from inertial loads when determine the friction needed at the pacer drive point.
A4. Dancer rollers do a great job of forgiving speed errors between pacer and follower sections. For intermediate tension zones, where the master and follower are controlled by rollers (with constant diameter) and a line speed reference signal is used to set the follower's baseline speed, a transducer roller is usually good enough to feedback web tension. However, on unwinds and winders, where diameters are changing, a dancer roller is a good alternative to forgive the more likely speed errors created by roll diameter uncertainty.

Lastly, for any process, I think it is a good idea to include a 'stall' mode in the process. In this mode, the line first pulls the web taut at zero speed before shifting to accel and run mode. The stall mode ensures a smooth transition from stopped to running conditions.

This all said, the slower you accel/decel, the less these issues are a big deal.

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 10:57 AM

Here's recent question:
Is it better to control tension and let strain fall where it may or control strain at whatever tension is necessary?

TJW Answer:
This is a big question, but let's try to answer it simply.
Most webs are elastic, so controlling tension or strain are really the same thing. If you have a web with an elastic modulus of 500,000 psi, then 0.5 lbs/ in / mil will be 500 psi stress or 0.1 percent elongation. It doesn't matter whether you run at 0.5 PLI per mil or 0.1 percent strain.

(All webs should be run at an average stress or strain that is 10-20 percent or less of their yield or break point.)

However, sometime things are not elastic.

For crepe paper, such as used to make masking tape, you can pull with too much strain and pull out the crepe, never to be recovered. This is a bad thing since the masking tape user is paying to get that strain for their conforming tape application.

In making nonwovens, before the binder is applied and cured or dried, the fibers can easily be pull out changing the product thickness, porosity, or other property. Overstraining the material is a bad thing.

Some webs are viscoelastic. If you pull on vinyl tape, it will stretch more every second or minute you keep the load on it. To avoid this, control strain, not tension. I was recently asked if web handling applied to cookie dough. Yes, but you better thing about the time - tension material property, the viscoelasticity of the dough.

Bottomline answer:
For most product, controlling tension, the easier to measure property is normal and let strain fall where it may.
It's still good to know what your typical web strain, because as I wrote my recent article in PFFC, strain is the secret to web handling. Strain determine specs for roller alignment, diameter variations, speed variations, etc.

For those special products that yield easily or are highly viscoelastic, look at controlling strain.

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 06:29 AM

Here is a recent question and reply on upgrading (or downgrading) a slitter tension control system.

Product: Small rolls of transfer ink on PET. Core diam 1-in., final diam <2-in..

We have an Allen Bradley MicroLogix PLC that we had to install some time ago to replace an old outdated touch screen system. We now have to replace an old outdate PID controller that was used to control the rewind tension or torque signal. I would prefer to use the existing PLC, but the integrator I brought in feels that an update time of 100msec would not be fast enough for rewind tension control. We could try it through software at a much cheaper cost instead of replacing more hardware, but I do not want to go down that path if it wil not work at all. The machine runs thin polyester film, 4.5 microns thick, 1-in cores, outside diameter is about 34mm, and machine speed is 450MPM. Any input articles, rules of thumb, etc. would be greatly appreciated.

(The dialogue posted in the first comment occurred via email, but the inquirer said he'd be happy to have it posted here for the opportunity for others to provide their input.)

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 06:42 AM

Registration isn't important to all converters. So many products are designed to have uniform properties in both MD (machine direction) and CD (crossweb direction).

Many converting process have a need for lateral registration. You need a gross registration to stay on your rollers, but you likely have a more challenging registration to hold a narrow coating or trim slitting margin. We usually don't call this registration, we call it web guiding (or chasing).

Registration usually involved pattern forming processes, whether printing, die cutting, or embossing. Each of these may be a stripe, needing only lateral registration, or full MD-TD registration of a 2-dimensional array of patterns.

For the live blog today, let's take questions in any issue of registration, including equipment, product, or process questions.

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 08:22 AM

How much should tension vary? This is a loaded question. Do you mean average tension as displayed by your tension transducer roller? ...or do you mean the actual tension in the web at point X.

I group tension problems into five categories.

1. Tension varies over time
2. Tension is too high
3. Tension is too low
4. Tension varies through the line (machine direction)
5. Tension varies across the web's width

Tension will vary. Even in closed-loop tension control, the web tension will oscillate around the set point by some percent.

Let talk about sources of tension variation, how they create process waste, and the proven remedies.

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 02:09 PM

No matter the topic, someone will always ask "What's the most important concept?" Whew. How do you boil everything you know down to one thing?

My dad once gave me advice on business (and life). He said "Follow the money".
In web handling, my advice is "follow the strain".

For web handling, winding, and slitting, it's all about STRAIN. What changes the dimensions of a product? Webs elongate in response to relative speed changes in drive rollers. Webs stretch differential to conform to misaligned and non-cylindrical rollers. Webs are forced into small gaps of nipped rollers or knives. Webs stresses relax within wound rolls as layers shift towards the core. (Continued)

Continue reading"Key Concept #1: STRAIN"

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 01:02 PM

A recent post in reply to the 'Two Pacers is One Too Many' post asked a pacer-related questions.

In a lamination process that has a reverse roll coater - drying oven - lamination nip rolls. Both the reverse roll coater and the laminating nip are driven with their own drives. We have been using a ratio control between these drives to maintain the web tension in the oven. The ratio control is being replaced with a closed loop automatic tension control system. Based on your experience should the reverse roll coater or the laminating nip be the pacer for this process? (Cont'd...)

Continue reading"Pacer Selection: Coater or Laminator?"

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 01:11 PM

All web processes should have a pacer. A pacer is a speed controlled section that drives a roller (and hopefully the web) to the process speed set point. The pacer doesn't care about web tension. In a slitter-rewinder, it's quite common to have a pacing section of driven rollers in the slitting and transport section and torque controlled unwind and rewinds.
If a process has two driven sections, say a two nipped pull roller stations, then one is the pacer and the other should be a follower. The follower pull roller drive can be controlled in draw control (sometimes called ratio control), in torque control (like a clutched or braked roller or a torque controlled motor), or in tension feedback / closed-loop speed control.
Occasionally, I run across a new web process that has two pacers - two speed driven sections ignoring web tension and just working to satisfy the speed set point. I'm not sure this should be thought of as two pacers, but a pacer and follower in 1:1 draw or ratio control. Is this a good idea? Not usually.

Continue reading"Two Pacers is One Too Many"

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 11:58 AM

John Bastarache wrote in with an interesting question:

"I have a question about tension in the sheet. If a sheet is 100 inches wide and trimmed down to 80 inches wide, is there more tension in the sheet at the reduced cross direction of 80 inches. Keeping in mind, that the draw is a constant."

If the tension was created by a dancer roller or applied torque, then the tension would increase on a PLI (pounds per inch of width) or psi basis.

However, you've included that the tension is created by draw. If you read my last two columns from a pffc-online.com, you'll learn the fundamentals of draw control. Draw creates tension through the web's material property, modulus, relating stretch (or strain) to stress (or tension).

Tension (lbs/in) = strain x thickness (in) x modulus (psi)

Therefore, the tension created by stretching (or drawing) on a force per width basis is independent of web width. However, if you are wondering about the total tension force (lbs), this will go up with width.

Force of Tension (lbs) = strain x thickness x width x modulus.

You can think of the t*w*E (thickness x width x modulus) as the web's spring constant.

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 02:46 PM

Draw control has always been a favorite topic of mine. Why? Because it is widely used in converting equipment, it is so simple to design, but the web's reaction to draw is somehow counter-intuitive.

I was recently working on a multi-station printing press. Presses are commonly run in draw control to keep the printing cylinder in registration. If you had each print station changing speed in closed loop tension control, you would hit your mark on average, but be off registration in both directions.

The tension in a press is based on draw control. The tension into the first press station is controlled by a constant tension infeed section (I hope it is). The tension between all subsequent stations is the baseline web strain modified up or down by the draw (relative speed) of each station to the first station. As a widely used rule of thumb, the series of print stations should be set at equal speed or a slight increase in speed from station to station. These guidelines imply that shifting to a lower speed (a.k.a. negative draw) would be a bad thing. Is it?

Continue reading"Don't be afraid of negative draw"

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 10:34 AM

I've been at a couple of client operations lately that have winding tension control in either open-loop torque control or dancer feedback to torque control using motors in low speed processes (low speed is < 300 fpm, semi-low speed is motor rpm is >2x core or roller fpm).

The main problem with torque control is "what torque gets to the web"? Sure, you can get a good torque proportional to amps out of the motor, but to deliver this to the web and create tension, you usually have to go through a gearbox and other sources of torque losses before you get to the web. I've learned (though I'm open to being taught by new technology, etc.) that better closed-loop tension control is found by using your tranducer or dancer roller to trim motor speed.

Continue reading"Is Motor Torque Control EVER The Best Choice?"

Filed under: Tensioning, Web Mechanics --- Tim Walker @ 03:12 PM

In preparing for my seminar at the CMM show this month, I was working up the list of what creates tension variations. Most people focus on machine direction tension variations without giving much thought to crossweb variations.

Why are crossweb tension variations snubbed? Probably several reasons. Unless the web goes slack, you many not see a crossweb tension variation. It rare and difficult to measure crossweb variations beyond slackness or left-right transducers.

What creates the biggest crossweb variations?

Continue reading"Crossweb tension variations"