|
Crimp force monitors (CFMs) continue to be a very useful tool in our industry for quality verification. They can save considerable amounts of time and money in terms of applicator tooling and material scrap, not to mention saving us from upset customers. They are typically well worth the investment in the proper environment and for those who fully understand them.
In the 15 or so years that I’ve been in the industry, CFMs have come a long way. They are much more accurate and, in many ways, easier to use than in the past. Some CFM manufacturers focus on user-friendliness but this tends to make the device less flexibility. In contrast, when manufacturers want to give the user a larger degree of control for a greater range of applications, devices tend to be more complicated to work with. However, in spite of these improvements, there are still considerable frustrations over using them.
Many people in this category still have many misconceptions about CFMs. Many believe that CFMs will eliminate their quality problems or they think they can simply put a CFM on any press. Some will ask the unqualified question, “Can the CFM detect one strand out?” However, this can only be answered after a series of questions about the application.
Four years ago, I wrote an article about the many misunderstandings and misconceptions about using CFMs. There are many benefits to the proper implementation of CFMs, but many users become frustrated because of a lack of understanding. This is why we still see machines and presses with CFMs that have been turned off.
Customers are still asking the same questions so I thought it would be helpful to address the topic again. I will discuss much of what was in the last article but I will expand on it. Most of the following information was presented at the Wire Processing Expo in May of 2008.
The intent of the article is to improve the knowledge of basic concepts for novice users of CFMs or those are thinking about implementing it as a part of their quality regimen. For those customers who have been using CFMs for many years, you might find this article to be basic; however, it’s never a bad thing to review the basics one more time. Individual features, technologies, or algorithms of the various CFMs on the market will not be addressed.
CFMs are more commonly implemented on automatic machines than on bench top because of the rate of termination and the inability for immediate inspection after the crimp. The primary concern is for those who want to implement CFMs on automatic cut, strip, and terminate machines. However, the same principles can be applied to bench-top applications.
To ensure that all bases were covered thoroughly, I consulted my colleagues in the industry; many of whom have been working with CFMs for longer than I. Input was gathered from Schleuniger Inc., Komax USA, OES, SLE USA and KMF Messtechnik GmbH. Therefore, you can feel confident that these concepts are not coming from just one source but from the all of the leaders in the field of crimp force monitoring. I formulated a few key questions to ask my colleagues.
1. Do most customers understand how CFMs really work?
The general consensus of the group was that most customers have an understanding of the basic settings and functions but not how it really works. Many customers believe that “CFMs will solve my quality issues” regardless of their process or equipment.
2. What is the most difficult concept to get across to customers?
Many customers do not understand that CFMs look at “process variation” and that there are many variables in the process. Subsequently, many customers don’t understand that different applications may require different CFM parameters. Lastly, interpretation of the CFM feedback (i.e. curves) can be quite challenging as well.
3. Where do you see the greatest problems when it comes to CFMs?
Most responses pointed to the quality of the materials and applicators, the set-up of the applications and the maintenance of the machines and applicators. However, many problems are caused because of the customers’ lack of knowledge. They don’t understand the CFM or how to use them most effectively.
4. What are the most critical components to ensure a successful CFM application?
The wire and terminal combination, and subsequently the headroom, have to be correct. The applicators have to be in good condition and the materials have to be good quality. Lastly, for new users, there must be a clear “champion” for the CFMs. Someone who will take the time to fully understand how to best use them and all the factors that need to be considered.
Basic Construction & Concepts
In the event that one of the readers is not familiar with CFMs, I wanted to cover basic construction and concepts. All CFMs are designed basically the same. There is a load cell that measures force, some type of triggering device to tell the CFM when to start reading and a control unit that performs the analysis.
Force Sensors
There are numerous types of sensors that will be placed either on the frame of the press (1), in the press ram (2) or in the base plate (3). They all do the same job; namely, register the forces that occur during the crimp. Some are better than others but the type and location will be determined by the manufacturer preference, the application or the price.
Frame sensors (Fig 2) are common today because they are less expensive and much easier to install on a press. They can be easily mounted by drilling and tapping a hole and screwing it in place. They also are very well suited for the majority of applications. However, they are not as sensitive as ring sensors. Because they are mounted on the frame, they see the force indirectly through the deflection of the press. For more sensitive applications, ring sensors (Fig 3) can be placed in either the ram or the base plate (shown), which is in the direct line of the force. However, using ring sensors is more costly because the sensors themselves are more costly plus they require custom parts to accommodate them.
 
Fig 2: Frame Sensor Fig 3: Sensor in Base
Typically, the only time you need to be concerned about the sensor type and location is when you are working with very small wires and terminals. For these applications, frame sensors will usually not work because they are not sensitive enough and a ring sensor will be needed in either the ram or the base. However, as you will read, wire size is not the only factor.
To demonstrate how sensitive CFMs can be, I will sometimes show a CFM pick out the difference between a white and black insulation. I’ve done this using two wires that are the same in every way except for the color. When the opposite color is crimped in the terminal, the CFM sees it as a bad crimp because the force signature is different. This should give you an idea of how sensitive these devices can be.
Triggering Device
A triggering device is required to tell the CFM when to start and stop analyzing the force signal. There are many types of triggering devices from proximity switches and light barriers mounted on the body of the press to encoders connected the shaft of the press. Usually the most reliable and accurate is the encoder, but they are also the most expensive solution.
Electronics
The signals from the sensors and triggers are fed into the electronics and combined in the control unit to create a force-angle or a force-time curve (Fig 4) consisting of a few hundred data points along the curve. Complex algorithms analyze the curve for shape and amplitude of each point on the curve and compare it to a known, good reference curve. Different manufacturers use different algorithms but all have a series of parameters that must be entered by the user. Some manufactures take a more user-friendly approach that is not as flexible. Others want to be more flexible but these tend to be more complicated. Unfortunately, in some applications, knowing what parameters to use can be tricky.
 Fig 4
Method of Use
Implementation of CFMs is basically all the same whether they are on a bench top press or on an automatic machine.
First: Set up the application
The crimp needs verified for all specifications (i.e. crimp height, bell mouth, brush, etc.) before proceeding with the “Teach-In” or “Learning” of the CFM. This may seem obvious but some customers make changes to the crimp after the learn process and they try to run production. Most often this just results in scrap because the CFM sees a different crimp curve and assumes it is bad. Therefore, make sure the application is perfect before beginning the Teach-in.
Second: Teach what is “Good”
Before the CFM can begin to evaluate anything, it needs to know what it is looking for. Therefore, the user must “teach” the CFM was is good. Typically teaching a CFM takes between 1 and 6 crimps regardless of whose CFM is used. These parts may be considered good production parts or they may be considered as scrap depending on the customer’s production practices. It is very important that the operator check these parts to verify that they are, in fact, all good pieces. If not, the CFM’s data may be faulty.
Keep in mind that the operator must tell the CFM what is good. If the operator runs the learn process with a crimp height that is too high and if the operator verifies that these are correct, the CFM will not know otherwise. It will look for, and accept parts with high crimp heights and it may reject parts with proper crimp heights. It is just like a computer. Bad data in equals bad results out.
If the learn process is successful, the CFM generates the reference crimp curve to which the CFM will compare all subsequent crimps. Each subsequent crimp is analyzed in shape and amplitude to determine if it passes or fails according to the tolerance parameters.
Third: Tolerance parameters
Tolerance parameters are the most complicated part about CFMs. If they are set too low, then good parts may be identified as bad crimps and the machine will stop often. This only frustrates operators and wastes time and materials. If they are set too high, bad crimps may pass as good. As we all know, this will only frustrate your customers which can be much worse than additional scrap.
For many applications, a default set of parameters might be used but the user needs to understand that not all applications are created equal. Shown below (Fig 5) are examples of two crimp curves for a 26 AWG (left) application and a 16AWG application. You can see that the basic shapes are similar but there are many differences in both timing and amplitude.
  Fig 5
Another example is the difference between running a small, gold-plated contact versus a larger, phos-bronze contact. With all other things equal, harder material will give more consistent forces. Gold is very soft and will show much more variation. Therefore, tolerance parameters may need to be slightly larger to allow for more variation.
+/- Tolerances: Areas & Shape:
The curves can be analyzed in a number of ways. The most typical are to look at the area under the curve and the shape of the curve. Both are usually monitored because it might happen where one of the parameters might be within tolerance when the other is out. Typical tolerances are around ± 4% for normal applications.
 Fig 6
Area of Analysis:
Typically, only part of the curve is used for the true analysis. The reason is that some applications will introduce “noise” into the curve that is not critical to the quality of the crimp. Either the CFM algorithms or a “filter” might be used to isolate what portion of the curve is used.
An example is the small peak(s) that occur in the beginning of the curve (Fig 5). This occurs when the tooling making the initial contact with the terminal and begins to bend the crimp wings over. Once they start to bend the force drops. However, the real part of the crimp has not even begun to occur yet so this part of the crimp is not critical. Also, there is usually much more variation in this part of the curve. The example shown (Fig 7) is an odd case but it might occur. Therefore, a filter might be used to eliminate this portion of the crimp from the analysis.
 Fig 7
Force Curve Zones:
Sometimes it is necessary to analyze different “Zones” (Fig 8) of the crimp to catch finer defects. Typically, in order to use these zones effectively, the rest of the process has to be very stable. However, each zone will have its own tolerance parameters.
 Fig 8
Choosing Optimal Parameters:
So, the CFMs are looking at a number of factors with each curve and the analysis involves numerous parameters. Although it may seem quite complex (and in a way it is), choosing the optimal parameters for your applications and work environment will lead to content operators and happy customers. Given that the materials and tooling are consistent and in good condition, a standard set of parameters can be used for most applications and the user should not need to worry about them too often. It is only for problematic applications when the user needs to go in and start experimenting with tolerance parameters. Regardless, the user should thoroughly understand how the CFM is analyzing the curve and what factors need to be considered.
Fourth: Run Production
Once the teach-in process has been completed and tolerance parameters are chosen, then production can be run. All subsequent crimps will be compared to the reference curve with the tolerance parameters chosen.
The Complete System
Before any CFM system can be used, the process or system has to be stable. When I refer to “the system,” I am talking about all of the factors that come into play when using CFMs.
When there is a challenging situation involving a CFM, most will only consider the wire, the terminal and the resulting crimp. The crimp might appear to be fine but the CFM has identified it as defective. Many other quality metrics (e.g. crimp height, crimp width, brush length, etc.) only involve one parameter. However, there are many more factors to consider with CFMs. The user needs to consider the terminals, the wire, the head room of the application, the applicator, the press and the operator or machine and finally the tolerance parameters of the CFM. Each of these variables can affect the resulting crimp curve and all play a part in the resulting forces that the CFM inevitably “sees.” Unfortunately, the CFM can not isolate specific variable(s) to analyze. In other words, it can not pay attention to some and ignore others and sees them all as a whole. Therefore, the entire system must yield consistent forces in order for the CFM to work properly.
Materials: Wire & Terminals
Not all materials are created equal, and usually, with less cost comes lower quality. However, there may be a point where paying less for the materials may cost you more on the production floor. Material quality must be consistent.
Terminals:
There are a number of factors that contribute to terminal quality. For instance, variations in material stock thickness will cause variation that the CFMs might detect. Variations in material stock thickness are to be expected to a degree and are usually not the main culprit. But, it is easy to imagine how these variations, if extreme, will adversely affect the ability for the CFM to do its job correctly.
I have also witnessed cases where a customer used a lower cost terminal that did not perform well in the applicator. The applicator was working as it should be, but the positioning of the terminals over the anvil was inconsistent because of the terminal quality. A similar terminal from a different manufacturer, although slightly more expensive, proved much more consistent and solved the problem.
Terminal material may also play a role in how much variation the CFM sees. Gold contacts typically show more variation than the same contact in another material. The reason is that gold is a softer metal, and softer materials will exhibit greater variation in forces. This is also the reason why CFMs can not be used on most applications involving pre-insulated terminals. The plastic insulation is too soft and exhibits too much variation.
Using oil on contacts also adds another variable. Although oil doesn’t cause too many problems when production is running at a normal pace, operators might see a higher rate of errors immediately after returning from a break. This is because the oil on the terminals between the anvil and the oiler has dried slightly. Therefore, the forces will be different. This might also give you an indication of how sensitive CFMs can be.
Improper care of terminals on the spools is another culprit of problems. The way in which the terminals are stored on the spool will affect the way in which the terminals are presented to the applicator. If terminals enter the applicator at odd angles, the crimp forces can be affected. In the example below (Fig 9), the terminals have not been well cared for. The different angle will cause variations in forces. Terminals entering the applicator properly can improve positioning over the anvil and terminal feed.
  
Fig 9
Similarly, light, side-feed terminals may also get angled slightly during the feeding process if the track is not adjusted properly. Once again, you might be in a situation where the crimp looks fine but the way in which it was put on the wire (i.e. straight vs. angled terminal) was different. Therefore, the CFM may identify it incorrectly.
Wire:
Non-concentric wire, as most of us know, will lead to stripping issues. Also, some insulation materials will adhere to the strands and cause stripping problems. If the insulation concentricity or adhesion is not consistent, a problem may be even harder to isolate. However, the CFM can detect variations in the force curve when strands have been nicked or cut easier than we will see the problem with the naked eye. These errors frequently can’t be seen after the crimp has occurred.
The number of strands in a wire also points to the question of whether the CFM can detect one strand out or not. One strand in a 7-strand wire will have a much larger impact on the force of a crimp than one strand of a 41-strand wire. So, if the CFM can see one strand out of a 7-strand wire, 2 or 3 strands may need to be out for a 41-strand wire.
Wire & Terminal Combination:
We all know this is not a perfect world. We know that there are cases where your customer has specified a terminal that is a little too large for the wire. For instance, it will be more difficult to monitor a 24AWG wire crimped into a terminal that is rated for 24AWG to 20AWG, than it is to monitor the same wire crimped into a similar terminal rated for 24AWG to 28AWG.
However, when the wire is small in relation to the terminal wire placement can be a critical issue. The operator may only see that the terminal is crimped on the end of the wire, but the CFM may be seeing significantly different forces. Below (Fig 10) are cross-sectional pictures of two consecutive crimps in which the wire is undersized for the terminal. The strands of the wire end up in different areas of the crimped terminal which may result in different forces. This is a case where it might be difficult to use a CFM.
  Fig 10
In general, traditional CFMs are most affective for applications of 24AWG and larger. Smaller applications can be difficult. Many of the factors I will discuss in the coming sections play a part, but the primary reason is that the forces related to just crimping the terminal onto the wire are too low compared to the other forces involved. In some cases, it is possible to detect 26AWG applications, but the smaller the application the more important it is to have a good head room and an applicator that is in good condition.
Head Room
The head room of a crimp is the difference in crimping forces when the wire is present and when the wire is not present. This concept plays a large part in answering the question, “Can the CFM detect one strand out.”
Below (Fig 11) is an example of a 16AWG wire application. The difference in force with and without the wire is approximately 47%. Therefore, roughly speaking, each strand of a 7 strand wire is will contribute about 6.7% of the force. If it were a 19 strand wire, each strand would contribute roughly 2.5%. It is not exactly like this but it is a decent approximation. However, if you are using tolerance parameters of ± 4%, you should pick up one strand out on a 7 strand wire but not on a 19 strand wire.
 Fig 11
In the example below (Fig 12), the peak force of the curve drops by only 26% so the affect of the wire on the overall force of the crimp is not nearly as much. In this case, each strand of a 7-strand wire will affect the force by roughly 3.7%; only 1.4% for a 19-strand wire. It is easy to see that if we use the same tolerances of ± 4%, the CFM will probably not see a crimp with one strand out as a defect.
 Fig 12
Some applications have a head room of 8% or 10%. These applications will be very difficult to work with because the majority of the force is just to crimp the terminal.
Applicators
Applicator quality plays a very big role in CFM effectiveness. An applicator that is in bad condition can introduce variation that the CFM will see.
For example, I ran tests using two different applicators for the same wire and terminal on the same automatic machine. The wire was a 16AWG bare copper and the terminal was a rear-feed, brass quick-disconnect; crimp heights and widths were identical. You can see below that the resulting crimps from the two applicators are very similar. However, the older applicator yielded a Cpk value of 0.65 and the newer applicator yielded a Cpk of 1.20. Although both values are not good, there is clearly a difference considering the same wire and terminals were used.
Figure 13 shows the crimp results from the new applicator that yielded a Cpk of 1.20. Figure 14 shows the crimp results from the older applicator that yielded a Cpk of 0.65. There is more consistency in the crimp in figure 13 than in figure 14. The CFM will be able to detect more of what is happening in the crimp itself because the applicator is not introducing additional variation.
  
Fig 13
  
Fig 14
Although the crimp may look fine from the outside, the CFM can see defects because the forces are varying. The biggest contributors to this problem are applicator age and lack of proper maintenance. Over time, applicators will wear out. Noise on the crimp curve can be introduced by any of the following: a ram that does not slide smoothly, worn tooling, inconsistent feed or inconsistent bell-mouth position. These issues might not be perceptible by a quick glance at the resulting crimp, but the CFM will see variation.
Sometimes an applicator has to “settle” after an adjustment or after new tooling has been installed. After the teach-in process, the forces may continue to drop. Therefore, the CFM quickly starts to register bad parts when they seem good. The reason is that the applicator needed to settle in to the new adjustment. Once it settles in force measurements will be consistent but it can be frustrating to operators.
CFMs can also help protect your applicator investment. Sometimes it doesn’t take many missed crimps to crack a die or anvil. The first missed crimp that gets stuck in the die will not affect the tooling. However, on automatic machines without CFMs one missed crimp can quickly become 5 or 6. CFMs will see these significant variations in force and might stop the machine before any tooling is damaged. Depending on the circumstances they might quickly pay for themselves in tooling cost savings.
The best solution for this is a regular maintenance plan for your applicators. I strongly recommend that anyone considering implementing CFMs should consider the age and quality of their tooling. This is especially true if purchasing a new piece of automatic equipment regardless of the brand. Putting old, worn-out applicator on a new machine is like putting old tires on a new Corvette. You simply won’t be able to get the optimal performance from the machine.
Presses
In order to use a CFM, the press has to have consistency in speed and shut height as well as be very rigid. The primary concern is with the older presses that many customers use. Many of the older presses are not rigid enough for use with a CFM. However, presses manufactured in the last 5 to 10 years are typically fine, provided they are in good condition.
Operator/Machine
In both cases, the key is wire placement. Inconsistent wire placement can cause problems with force curves in some applications. This is especially true of the depth of insertion. Novice operators or machines in poor condition may have issues with wire position.
Conclusion
When used properly, CFMs can be a tremendous asset on any production floor. They can save you considerable amounts of money in tooling and scrap costs. They might also allow you to use lower skilled labor on certain applications. CFMs will help you better understand the process and might even identify some problem areas. This might be seen as either a pro or a con depending on the situation. Because the CFMs look at the complete system, using them might require you to improve quality in other areas like applicator maintenance or material quality. Finally, it is something you can sell to your customers. Quality monitoring is always seen as a positive.
However all of the factors need to be considered. For new users, there can be long learning curve. It is very important that there be at least one key person and that this person be thoroughly trained. Proper training is extremely important. They must also be willing to take the time to really understand the best way to utilize the CFMs.
Users must understand that there are many components to the system and that the CFM can not analyze certain variables and ignore others. CFMs will look at variation of the entire process, which includes the wire, terminals, applicators, operators and machines. Because of this, not all applications are considered equal. Make sure that your equipment is well maintained and that you are getting consistent quality from your materials.
I hope this information has shed some light on the different factors and challenges of using CFMs on your production floor. Although there can be challenges, when used properly, there are many positive benefits of using CFMs for those who understand them thoroughly.
I would like to thank the following individuals for their help in writing this article: Verena Behrmann, KMF Messtechnik GmbH; Michael Reeve; OES Inc., Erich Moeri, Komax USA; Enrique Duarte, SLE USA.
|