You are here: Products & Solutions / Roll Forming / Edition December/2009
Friday, January 22, 2021

Virtual roll forming is already a reality

Tube and section machining
08 / 11 / 2008 - Author: Dietmar Kuhn

Walter Frick of interviewed managing director Stefan Freitag.

The profitability of roll forming sinks or swims depending on your experience and knowledge - either that in the heads of designers and machine operators or in software packages for design and manufacture. So what is the role of software in the roll forming pro-cess? We spoke about that to Stefan Freitag, managing director of data M Sheet Metal Solutions.

Stefan Freitag, managing directorof data M in Valley, Germany

Blechnet: Stefan, how does the right software improve the advantages of roll forming even more?

Freitag: With the design software for cold rolled sections obtainable on the market today it’s possible to create new designs even of complex sections in a matter of days. Instead of the six weeks it might have taken you 20 years ago. Analytical simulation tools enable the designer to check for flaws in the layout phase already. They support you in particular in detecting linear expansion. Because of the roll forming process this is one of the main causes of defects on the finished section. But calculations of this kind always base on the designed cross-sections. Spring-back between stations or compliance of shafts – as with high-strength material for example – doesn’t enter into it.

You can only do this with an FEM calculation. But the use of software has still made design operations much more efficient. Today you can very easily derive similar section cross-sections from ready existing designs because the basic calculations for the flower are fully parametric. It doesn’t matter if the shape or the geometric dimensions including sheet thickness are different. Every change made to the final cross-section is automatically adopted in the roll forming flower. That makes for effective use of roll forming experience that’s already accumulated. The systems are also transparent, meaning that a designer is able at any time to reconstruct the calculation of bandwidth or method of calibration that’s been used.

Additional calculations like statistical strength or theoretical spring-back help the engineer to optimize their first design in as much as possible. The software also supports the design of roll tools. Specially attuned functions can be used to automate certain design processes. These include the implementation of forming rolls or definition of clearance angles. Manufacturing data are automatically extracted from the designed rolls. That also means a substantial time-saving because it does away with all the detailing effort.

You can also search for and re-use rolls saved in a database. That economizes on manufacturing time and cost. A further important point is the greater use of accessory rolls. As cross-sections become increasingly complex a profiler today can no longer manage just with the classic station arrangement of top and bottom roll plus side roll.

Blechnet: What influence does the software have on the machined result?

Freitag: Today you find very complex cross-sections being manufactured by cold roll forming. The quality achieved is also a result of the software that’s used. Calculations become reproducible and the results reapplicable. Typing errors are ruled out when calculating the neutral fiber or configuring the flower. The use of software also gives you the accuracy of NC lathes that’s possible today, which is another contribution to quality. You mustn’t forget that the quality of the end-product depends on many process factors – the quality of the design, of the material, of the roll production, of the rolling plant and of the setting up must all be good too.

If one of these factors doesn’t line up properly you’ll have problems with the end-product.

Blechnet: Can you simulate the roll forming process by software entirely, is virtual roll forming possible?

Freitag: Yes, virtual roll forming front to end is already a reality. When it emerges from design the set of rolls is automatically dressed for simulation. The mesh necessary for the strip is automatically extracted from the position of the bending points, the material and process parameters are defined. In other words are you manufacturing from a coil or blanks, should cutting of the strip to length be simulated, do you want to simulate with the simplified model, or must you calculate with rotating rolls and friction.

If it’s a case of ready punched strip the hole pattern can be transferred to the mesh for the material. This shows whether the geometry or position of the holes alters. Based on this information it’s possible to correct the shape of the, in most cases, very costly punching die as early as the drafting phase.

That in turn saves time and cost when starting up. The use of drawing dies or molds in simulation isn’t a problem either. We’ve worked on several hundred industrial projects in recent years that in most cases involved optimizing sets of rolls by means of FEA.

Feedback from these projects attest to very good harmony between theory, or simulation, and practice. There are some instances where the results don’t match, and then we have to determine the cause. With high-strength materials it’s often compliance of shafts that produces a different result on a machine to what was simulated.

So you use the simulation results to calculate the compliance of the shafts, set the roll gap appropriately, and calculate the profile anew. The larger roll gap means that the material is not formed more than necessary, and you can then achieve extremely small spring-back rates.

If the conditions accompanying theory and practice agree, the result will be right too. Of course simulation always proceeds from the figures that are issued by design. So if the tolerances of the produced rolls are not correct or the rolls are wrongly produced in the first place – that’s still possible even in the age of numerical control – the results of simulation aren’t what you end up with in reality.
In such cases the COPRA® RollScanner is a solution for testing rolls.


A veritable treasure trove of rolls

Tube and section machining: CAD / CAM
11 / 24 / 2008 - Author: Michael Corban

Michael Corban of interviewed managing director Albert Sedlmaier.

Sets of rolls frequently go into storage once they have been used. Assuming 10,000 tools you may soon have rolls worth a million or more lying unused in a warehouse. A front-to-end process chain from data M Sheet Metal Solutions in Valley, Germany is aimed at scanning, saving and retrieving the geometry data for access to this treasure trove, as managing director Albert Sedlmaier explains.

Albert Sedlmaier, managing director of data M in Valley

Blechnet: Albert, why is it worth re-using roll tools?

Sedlmaier: Big companies will manufacture about a hundred new sections in the course of a year. An average set of rolls consists of between 200 and 250, which can fast add up to a few tens of thousands before the year’s out. Some of these new tools are bound to be already available in stock – but they just can’t be traced. We reckon that a good 10% of the rolls can be found in the inventory. That makes for a simple calculation. For a hundred new sets each of about 200 rolls I’m already stocking 2000 of them. Since each roll costs an average of 150 euros, I have potential tooling worth about 300,000 euros just waiting. I only need to find the right tools. If new rolls are added, the proportion will look even better longterm.

Blechnet: Why don’t companies use as many used roll tools as they could?

Sedlmaier: Even if users attempt to restrict themselves to certain basic forms, there are still a whole number of project- or profile-related rolls that aren’t so easy to re-use. Often you only have paper drawings of the major applicable tools, and it’s impossible to search through them all.

Blechnet: Do CAD data make it easier to retrieve them?

Sedlmaier: Even that doesn’t help because roll tools are operated, and possibly remachined. In most cases nobody does the updating for this – meaning that the CAD data aren’t a reliable picture of the geometry of the particular roll. Then there’s a second aspect. Tools that have rolled some tens of thousands, sometimes perhaps millions of meters of material show wear and tear. The profile is no longer that of the original design. Even if I did find matching CAD data, I still don’t know what exactly the roll looks like.

Blechnet: So what should someone do so they can re-use old rolls?

Even if you have the CAD data of the tools, you still do not know their precise contour. The roll scanner can detect how worn they in fact are. These data automatically land in the recycling database, allowing a targeted search for old tools.

Sedlmaier: The simplest way is to create a process chain, as we do by combining our COPRA® Recycling Database with the COPRA® RollScanner. The process chain then looks as follows. A roll is scanned and the real contour that’s found is transferred direct into the recycling database. Parallel to that we break down the contour into Fourier coefficients, which is useful for the search algorithm. In this way the user can speedily and simply search for similar rolls in their specific recycling database.

Blechnet: How does the search work?

Sedlmaier: We work in multiple steps because we obviously want a result as fast as possible. First of all the roll borehole must match, then of course the maximum diameter and maximum width. In that way I can fast restrict the search because it’s only a matter of geometric comparison. So there are only a small number of rolls that come into question for the intended application.

Another important criterion is that the rolls I find shouldn’t have an intersection with the contour I’m searching for. If the search doesn’t produce a result, a second approach is a similarity search by Fourier coefficients, enabling me to find similar contours fast. Here the contour of the recycling roll is analyzed and saved so only rolls are found that really have the required contour. There were approaches in the past but in Unix days neither harddisk capacity nor computing speed were adequate for obtaining a result in an acceptable time.

At about 150 euros per tool you can soon be stocking rolls worth millions. Costs are easily cut by re-using just 10% of them.

Blechnet: So how fast is the search today?

Sedlmaier: Normally a roll is found in a matter of seconds – and it has to be for such a solution to be accepted. Imagine I had to wait just one minute for each roll of a set of 250 tools. That would take half a workday, which is out of the question. But with the SQL databases available today under Windows obtaining a response within seconds is no longer a problem. Seamless integration of the individual steps as in our solution also promotes acceptance, especially because it simplifies operation.

Blechnet: What does the application look like for the designer?

Sedlmaier: Naturally they want to know what rolls already exist or what sections have already been designed. So first they sketch the rolls they’re looking for. If a matching recycling roll is then found it can be inserted in their CAD system in a separate layer.

What’s really clever about searching by Fourier coefficients is that the designer can specify the oversize for the roll they’re looking for. So even worn rolls can be used again. Also, only those rolls come into question that can be reworked within the set limits and with relatively little effort to produce the required contour fast.

The recycling roll that’s found is subsequently linked with the designed roll so that the production data, such as material and sawing lists, clearly show that this roll already exists and only needs to be remachined. Once that’s done, the found recycling roll is of course, and automatically, removed from the recycling database.

Blechnet: And when does the roll scanner enter the scene?

Sedlmaier: As a rule you work with two databases. One holds the tools that are in use, the current ones, while the other is the so-called recycling database.

Given the data, you could simply load the data set after use into the recycling database. But you still have to ask what contour the used tools really have. And I’ve no idea of the contours of older roll sets. That’s precisely why we developed the roll scanner. At one go, and automatically too, the data are suitably formatted for the database and saved, including the Fourier coefficients. From then on the geometries – and the real ones on the tool – are ready for retrieval by the user in the recycling database.

Blechnet: How much effort does this involve for the user?

Sedlmaier: The actual scanning operation is a fast matter. The main effort is one of logistics. In addition to transporting the rolls out of storage and back again, the organization of storage plays a major role. Because I can’t fetch the scanned tools from storage when I need them if I don’t identify and store them properly. The user doesn’t have to bother about managing the data sets. The COPRA® RollScanner is designed to take care of that. It gauges the roll tools speedily and accurately, saves the results in the database, and can issue a report direct with information about the quality of the rolls. Our development focused very much on ease of operation.

COPRA® RollScanner from data M