Material for this article supplied by: The LVD Corp., Plainville, CT
Press Brake BasicsThe "ABCs" of press-brake forming are changing from down-stroking/up-stroking, air-bending/bottom-bending, special-tooling/standard-tooling, etc. to Smart Bending. It is much more important knowing how your processes will perform, than what methodology you use.
Press-brake fabricating departments, to some degree, have defied just in time (JIT) and total quality management (TQM). "Economic" lot sizes are selected to amortize setup times, and not necessarily to reduce work in process. Jobs are routed to the next available press brake, with little regard to that machine's particular capabilities.
Press-brake forming is an art, and greatly dependent on operator skill. Therefore, it is expected that varying levels of quality and productivity are to be achieved by different operators. Consequently, inconsistent results are accepted and out-of-tolerance parts are reworked, scrapped or simply passed to the next stage. Excessive setup time adds no value to the bent part.
The first step toward finding a solution is identifying the problems. However, generalizations cannot be made, because of the diversity of applications for which a press brake is used. For example, batch quantities are instrumental in the productivity of a press brake. If two setup test blanks are wasted, for a production run of 100 parts, only two percent scrap is achieved. If two setup test blanks are wasted, in a production run of 10 parts, 20 percent scrap is achieved. Although the parts may be similar, the problems and solutions are totally different.
Other problems only appear to be elusive; however, they are very predictable. It is not unusual for the flat blanks to be incorrectly sized. This can be due to "bend allow-ance" -- calculating errors and gaging errors at the shear. When the part is formed, all of the flat blank size errors arrive in a single flange of the bent part. Which flange? The answer is always the same: It depends on the bend order and gage locations, which are usually selected by the operator.
NC and CNC controllers were developed as input/output devices in which the operator would key in machine coordinates (ram and back gage locations) and the controller then would position the axes according to the request.
As machines gain more programmable features (e.g., tool compensation, lateral part positioning, bending pressure and speed) Fig. 1, the controllers can be equipped with many mathematical formulas to predict machine coordinates, based on piece part geometry.
Fig. 1 -- Front gages, material
handling, and integrated tooling can be controlled by the part program.
The current "state of the art" in CNC controls is macro programming and data mining where redundant questions are eliminated and existing empirical data is extracted for future use using statistical and visualization techniques to present this knowledge in a form that is easy to understand, Fig. 2.
Tests on press-brake controllers have shown that not only the springback but even more the behavior of the material during the bending phase are determined by material properties requiring different depth settings of the punch to reach correct bend angles. The friction between the blank and die occurring during the bending process and actual bend radius achieved are other elements that are hard to define yet certainly affect the bend angle.
Fig. 2 -- The bending
environment is displayed in a form that is easy to understand, including instructions on
how to handle the part during production.
Most CNC machines use a formula (similar to the formula for a right triangle) to approximate the ram penetration to achieve the requested angle. This formula is based on theoretical calculations and ignores the material characteristics. Errors in measuring the exact tooling geometry also influence the trigonometric calculations.
Typically the press-brake operator is expected to learn how bad his machine performs, and manually compensate for all errors. For example, if a part is formed from 10 ft. of 10 ga. stainless steel and the requested angle was 90 deg., however, the angle produced is 95 deg., the machine operator can conclude that under these conditions he can achieve a 90-deg. bend in 10 ft. of 10 ga. stainless steel simply by requesting an 85 deg. angle. This practice is called "lying to the control." (Ask any CNC machine operator).
The current trend in press-brake controllers is to identify a potential problem, then automatically compensate for it. In other words the machine controller is emulating a good press-brake operator.
Fig. 3 -- "Data Mining" is the first step
in documenting/eliminating the variables in your forming process.
Databases are being used in some press-brake controllers to automatically store/retrieve correction data. When jobs are corrected the computer logs the corrections along with part parameters in a database, Fig. 3. When new jobs are entered, the computer searches the database for corrections that would pertain to this job, and applies them to the new program.
These built-in expert systems can be "transparent" to the user. The system automatically records "experience" information that is to be used in a later stage. All the operator notices is a gradual increase in the performance of the bending system. In actual practice, this means that, in time, almost no redundant external corrections will be required.
Material sensing also is used to improve results. An electronic micrometer can be interfaced with the machine controller. The operator passes the flat blank through an automatic thickness measuring device. The exact sheet thickness is transferred into the machine controller, and the punch position is calculated, based on the actual measured thickness.
Some systems use the press brake itself as a giant micrometer to measure the blank thickness. Others use load sensors to "feel" the hardness of the blank. The data obtained from blank sensing is used to speculate the ram reversal position, to achieve the requested angle.
An in-process angle-measuring system that "positions to actual angle" and not to a predetermined ram position now is available. This closed-loop system uses a digital measuring device to monitor the piece-part angle as it is being formed. Because the angle is monitored from the outside, the results are not affected by variations in material thickness. The "in-process" measured angle is the "steering" parameter for the ram position control.
Fig. 4 -- Actual angle
achieved is measured and corrected in process by a completely "closed loop"
system.
To achieve the requested angle, this type of net shape forming (NSF) system, Fig. 4, does not rely on the geometric tooling values or the material properties, which are in the majority of cases unknown to the operator. Corrections are applied as needed, while the part is being formed, to insure perfection without significant manual trial and error (teach/learn) bending.
Graphic simulation, Fig. 5, is used to verify the entire press-brake bending environment. Material handling, support arms, even tool shapes can affect the setup and run times. For example, knowing what tools can be used to form a certain part is useful in the planning process. If a part is easier to form using a gooseneck punch, however, the part could be formed with a standard straight punch, the decision whether or not to mandate a punch change for this job, can be based on the simulation results.
Fig. 5 -- Non-parallel flanges and
holes are displayed in the bending simulation to aid the programmer/operator.
Once the optimal bending sequence has been determined (automatically or manually), it can be simulated on the computer screen or printed on paper. During this simulation, the user can see all the details of the production set, (punch, die and back gage) allowing the operator to determine the proximity of the product to the tools at each stage of the process. In this way, it is possible to create a production document that shows potential dangers to the press-brake operator.
The material handling needed between the current and next step of the bending sequence is optimized and displayed on the screen.
Graphic systems also offer the capability to automate programming in an efficient manner. The user enters the shape of the required sheet metal product into the system and the bend program (sometimes referred to as the G-code) is automatically generated. This means the operator only needs to enter part shape, and not machine parameters. Without graphic simulation, the NC-program must be created manually. This means that the operator has to describe every step (sequence) of the production process. He must determine exactly where all the axes of the multi-axis press brake have to be positioned in every step of the production of a certain part. Besides that, he also must determine how the auxiliary functions have to be regulated. This can become a very cumbrous job and rob valuable machine time.
When three-dimensional part drawings are available as they will be cut by another machine (e.g. punch press or a laser cutting machine) they can be transferred to the press-brake programming station. The DXF (data exchange format) drawing should contain the required bending information: General information of the part, the bending lines with their angle value, radius value and bend allowance value.
When we review the evolution of the modern press brake, we see that the current hardware has been available for quite some time: 1963 up-stroke brake, 1969 precision-ground tooling, 1970 automatic back gaging, 1980 CNC-deflection compensating, 1984 servo-hydraulic, etc. These items are commonplace now, or at least their availability has been well known for many years.
Almost all of the innovations for press-brake forming have been to harness existing tools. Increased use of computers and press-brake forming software are the fundamental elements for achieving quality and productivity goals. MF