第一部分：DDR、PCIe 正向力分析
最近这段时间，我开始陆续用 Abaqus 做一些连接器端子的应力分析。目前接触得比较多的，主要是 DDR、PCIe 这类产品的正向力分析。严格来说，我现在还在持续摸索和积累阶段，还谈不上把这块完全吃透了。不过这段时间确实已经开始实际上手，也做了一些模型，踩了一些坑，对正向力分析这件事有了比以前更具体的理解。所以我想先写一个简单的总结，算是把最近这一阶段做的事情、一些基本认识，还有分析中碰到的问题，先记录下来。一方面是方便自己后面回头看，另一方面也是给自己这段时间的学习做个阶段性整理。
一、最近为什么开始做正向力分析
我本身之前一直是做连接器产品工程这条线，对产品结构、规格、可靠性、量产导入、客户问题这些会更熟一点。以前碰到端子正向力相关的问题，更多还是通过测试结果、产品经验、结构理解去判断。比如这个端子压下去够不够力，这个结构会不会太硬，哪个位置可能容易应力高、容易变形，这些以前其实也会有一些经验判断。但经验毕竟是经验，很多时候你知道“大概会有问题”，和你真正把这个问题在仿真里看出来，还是两回事。最近开始做 DDR、PCIe 这些产品的正向力分析，对我来说，一个很大的意义就是：以前很多东西是靠经验推，现在开始能试着用分析的方式把它更具体地看出来。说白了，就是从“凭感觉知道哪里不太对”，慢慢走向“我能看到它为什么不对，风险大概在哪”。
二、我现在理解的正向力分析，核心是在看什么
正向力分析，表面上看是在看一个力值，但实际上，它不是只看“这个端子压下去有多少克力”这么简单。我现在理解下来，它至少是在看这几个东西：
第一，目标行程下，端子的正向力够不够。这是最直接的，也是最基础的。
第二，端子压下去以后，应力是不是太高。因为有些结构为了把正向力做出来，会把弹性臂做得比较硬，力值虽然上去了，但高应力也跟着上去了。
第三，端子的变形方式是不是合理。有些模型最后反力数值看起来还行，但变形模式其实不太顺，局部会有不自然的扭曲、局部过度变形或者接触状态异常。
第四，后面回弹会不会有问题。也就是说，它压完之后还能不能比较健康地弹回来，会不会有永久变形风险。
所以我现在越来越觉得，正向力分析不是单纯看一个结果值，而是在判断这个端子整个受力状态是不是健康。
三、最近做 DDR、PCIe 正向力分析时的一些基本思路
最近我做的这些分析，虽然产品不完全一样，但整体思路其实差不多。一般来说，就是把端子主体建出来，然后设置刚体或者对应的受压对象，通过位移加载去模拟它被压下去的过程。再根据实际结构，去定义哪些位置是固定的，哪些位置是真实受约束的，最后通过接触和反力来提取结果。说起来很简单，但真正做的时候，我现在感受最深的还是两件事：
1. 边界条件一定要尽量贴近真实。这个东西我现在体会越来越深。因为端子在产品里不是“悬空自由弹”的，它一定是受到塑胶、固定脚、包胶区、装配状态这些条件限制的。如果边界条件设得不对，最后分析出来的力值和应力分布就很容易失真。所以现在我做模型，不会只想着“把它固定住就行”，而是会更注意：它在真实产品里，到底是怎么被限制的。
2. 接触问题往往比想象中更麻烦。正向力分析里，接触基本绕不开。压头和端子接触，或者端子和对应配合区域接触，这个过程本身就会带来很多非线性。这也是我最近做分析时最明显的一个感受：很多时候模型不是不会建，而是建出来以后不一定好算。尤其是接触初始状态不够顺、局部倒角太尖、摩擦设置不合适、网格质量一般的时候，模型就会变得很容易卡。
四、这段时间做下来，我主要关注哪些结果
最近我做这些 DDR、PCIe 正向力分析时，重点一般会看下面几类结果。
1. 力值结果。这个最直接。就是在目标位移下，看它的反力到底是多少，能不能满足产品需求。但现在我也越来越不会只盯着一个数值看了。因为有时候力值看起来够了，不代表这个设计真的就合理。
2. 高应力区域。这一块也很关键。像端子的根部、折弯过渡区、接触区附近、局部几何突变的位置，通常都是要重点看的地方。因为这些区域如果应力太高，后面就可能带来屈服、残余变形、寿命下降这些问题。尤其是 DDR、PCIe 这种产品，本身对接触稳定性要求就高，所以不能只把正向力做出来，还得看这个力是不是“健康地做出来的”。
3. 变形状态。这个东西我现在很重视。因为有些模型表面看，结果数值还凑合，但你一看变形云图，就会发现这个端子受力路径不太自然。这种“不顺”的感觉其实很重要。很多时候结构后面出问题，早期在变形模式上就已经有苗头了。
4. 回弹和残余风险。如果模型有继续做到卸载阶段，那我也会去关注它弹回来的状态。因为这一步更能看出端子的弹性储备到底够不够。有时候压下去时看着问题不大，但回弹阶段反而更容易暴露问题。这个也是我最近做分析时，感受比较深的一点。
五、最近做这类分析时碰到的一些典型问题
最近这段时间做下来，我觉得正向力分析虽然算是比较基础的应力分析类型，但真正上手以后，也并不轻松。
1. 收敛问题比较常见。这个应该算我最近碰到最多的问题之一。尤其是接触一建立起来以后，模型就容易开始不好算。比如：初始接触不够平顺、端子局部倒角太尖、摩擦影响比较大、网格不够理想、位移推进过程中局部接触状态变化太快，这些东西都会让求解过程变得比较痛苦。
2. 下压能过，回弹不一定能过。这个问题我也有比较明显的体会。有些模型在下压阶段还能顺利跑过去，但到了卸载回弹阶段，反而容易报错。因为到了回弹阶段，接触关系、应力释放、变形恢复这些行为会更复杂，模型的非线性也更强。所以很多时候，真正难的不是“压下去”，而是“压完以后怎么回来”。
3. 力值和高应力经常是矛盾的。这个其实也是连接器端子分析里很典型的一个问题。你想把正向力做大，通常就会想办法让结构更硬一点；但结构一硬，应力往往也会上去。所以很多时候分析不是在追求“最大力值”，而是在找一个更合适的平衡：力值够用，同时结构别太危险。我觉得这也是正向力分析真正有价值的地方。不是只告诉你“够不够”，而是帮你看“代价是什么”。
六、这段时间做下来，我对正向力分析的一些理解
最近做了 DDR、PCIe 这些产品的正向力分析以后，我自己有几个比较明确的感受。
1. 正向力分析不是只看一个数。以前没真正开始做的时候，可能会觉得：反正最后看一下正向力值，不就差不多了吗？但真正开始做以后会发现，不是这样的。真正有价值的，是你把力值、应力、变形、回弹这些东西连起来看。
2. 仿真结果一定要结合产品理解。Abaqus 不会自动替你理解产品。模型跑出来了，结果摆在那里，怎么解释，哪些是合理的，哪些是危险的，最终还是要靠对产品结构和实际工况的理解。所以我现在越来越觉得，自己以前做产品工程积累的那些经验，其实在这里很有用。因为你不是只会点软件，而是知道这个端子在真实产品里到底是怎么工作的。
3. CAE 的价值在于“提前看问题”。这个也是我现在最认同的一点。以前很多事情要等到样品、测试、异常出来以后才能更明确。而现在通过分析，至少可以先把一些明显风险前置暴露出来。虽然仿真不能完全代替测试，但它确实可以让设计判断更提前一点、更主动一点。
七、对我自己的意义
对我来说，最近开始做 DDR、PCIe 的正向力分析，不只是多学了一个分析案例。更重要的是，我感觉自己确实开始从产品工程往 CAE 这条路上真正走起来了。以前我的优势更多是在产品理解、规格理解、异常处理、客户问题应对这些方面。现在如果把应力分析这块慢慢补起来，那我后面看问题的维度就会更完整。简单说就是：以前我更多是靠经验处理问题；现在我开始尝试把经验和分析结合起来，去提前判断问题。我觉得这个变化还是挺重要的。
八、后面还要继续补的地方
现在这块我肯定还在继续积累，后面还有很多东西要补。比如：不同产品结构下正向力分析的差异、回弹和残余变形做得更扎实、插拔力分析继续补、分析结果和实际测试结果多做对比、接触/摩擦/收敛等核心难点继续啃。总的来说，最近这一阶段做 DDR、PCIe 的正向力分析，对我来说算是一个开始。还不算很深，但确实已经迈进来了，也开始形成一些自己的理解了。
结尾
这段时间把 Abaqus 正向力分析真正做起来以后，我最大的感觉就是：这件事表面上是在分析一个端子的力学表现，实际上是在训练自己怎么更早地看懂一个产品的风险。尤其是 DDR、PCIe 这类产品，本身对接触状态、结构稳定性都比较敏感，所以正向力分析还是很有意义的。我现在还在持续做，也还在持续踩坑。但至少这一步已经开始了。后面再继续积累下去，我希望自己不只是会做产品，也能越来越会用分析的方法，把产品看得更深一点、更早一点。
第二部分：Abaqus插拔力分析小结
最近这段时间，除了做一些 DDR、PCIe 的正向力分析，我也开始做连接器端子的插拔力分析。这类分析做下来，我的感觉很直接：插拔力分析比正向力分析更复杂，也更花时间。因为它不只是看端子压下去以后的反力，而是要模拟端子和插卡在整个插入、拔出过程中的接触、滑动和受力变化。这里面会牵扯到摩擦、倒角、接触稳定性、网格、分析步设置这些内容，所以对模型细节要求比较高。这篇就简单做个小结，记录一下我最近做插拔力分析的一些认识。
一、我现在怎么理解插拔力分析
插拔力分析，说白了，就是看端子和插卡在插入、拔出过程中，力是怎么变化的。它主要不是看一个最终结果值，而是看整个过程：什么时候开始接触、倒角过渡顺不顺、插入力峰值大不大、拔出时阻力大概多少、力曲线有没有异常波动、接触过程中有没有卡顿或者局部应力过高。所以这类分析会比正向力分析更接近真实使用状态，但同时也更难做。
二、插拔力分析里几个比较关键的点
1. 摩擦系数：摩擦系数对插拔力影响很大。因为插拔过程中，端子和插卡表面一直在滑动，摩擦本身就是插拔力的重要组成部分。这个参数如果设得太大，插入力可能明显偏高，模型也更容易不好收敛；如果设得太小，虽然可能更容易跑，但结果又可能偏离实际。
2. 分析步稳定性参数：插拔过程里，接触状态一直在变，所以分析步的稳定性设置很关键。特别是在刚接触倒角时、接触点切换时、局部滑动突然变强时，模型容易不稳定。自动稳定、时间增量这些设置，很多时候会直接影响模型能不能顺利跑下去。
3. 分析步数：插拔力分析不能走得太粗。如果步数太少，每一步跨得太大，容易出现接触变化跟不上、力曲线不平滑、局部峰值抓不出来等问题。尤其是端子过倒角的时候，这一段通常最敏感，所以步数一般要分得更细一些。
4. 插卡倒角：倒角如果不顺，或者太尖，模型里就很容易出现接触切入太生硬、插入力峰值过大、端子局部应力太高等问题。所以倒角不只是装配体验问题，对分析结果本身也很关键。
5. 网格细密程度：网格对接触和结果都很敏感。接触区、倒角过渡区、端子根部、小圆角附近这些地方如果网格太粗，接触不容易平顺，应力也可能不准。但全模型都做得很细，分析时间又会太长，一般还是关键区域细一点，其他区域适当放开。
三、插拔力分析的一般流程
我现在做这类分析，大概就是按下面这个思路走：先整理端子和插卡的几何 -> 建立材料、接触和边界条件 -> 设置插入过程的位移加载 -> 根据情况增加拔出过程 -> 调整分析步和稳定参数 -> 检查网格，重点细化接触区和高应力区 -> 跑结果后看力曲线、应力、变形和接触状态。整体上看，流程不算特别复杂，但真正难的是每一步的细节。
四、我现在常见的报错和原因判断
最近做下来，常见问题主要还是这些：
1. 接触后不收敛：可能是倒角太尖、初始接触不顺、摩擦过大、网格太差或步长太大。
2. 下压能过，继续插入就不行：说明接触路径开始复杂了，可能是某个位置卡住了、接触切换太突然、倒角或者路径不够顺、稳定参数偏弱。
3. 力曲线跳动大：可能是步数太少、网格太粗、接触不平滑或摩擦设置不合理。所以很多时候，问题不一定是模型完全错了，而是某个局部太敏感。
五、插拔力分析优化的几个常见方向
最近我自己的经验是，插拔力分析如果不好算，或者结果不理想，一般可以从下面几个方向去看：
1. 先看倒角和接触路径顺不顺。很多问题其实不是参数问题，而是产品本身接触路径不够友好。
2. 再看摩擦系数是不是合理。摩擦过高，经常会让模型又难算、力也偏大。
3. 调整分析步和增量控制。让过程更细一点，很多时候会更稳定。
4. 优化网格。特别是接触区、倒角区、根部这些位置，网格质量很关键。
5. 检查边界条件是不是太死或者太松。边界条件不合理，也会影响结果和收敛。
六、我的整体感受
总的来说，插拔力分析确实是一类比较“吃细节”的分析。摩擦系数、稳定性参数、分析步数、倒角、网格，这几项都会直接影响结果和求解过程。再加上插拔路径本身比较长，所以分析时间一般也比较久。但也正因为它复杂，它的价值才更明显。因为它能更接近真实插卡过程，也更容易把结构里的问题提前暴露出来。我现在这块还在继续摸索，但已经越来越能感觉到，这类分析不只是软件操作问题，更是产品理解、接触理解和建模细节的综合能力。
Part 1: DDR, PCIe normal force analysis
Recently, I have begun to use Abaqus to do some stress analysis of connector terminals. At present, I have been exposed to a lot, mainly the normal force analysis of products such as DDR and PCIe. Strictly speaking, I am still in the stage of continuous exploration and accumulation, and I cannot say that I have fully understood this area. However, during this period, I have really started to get my hands dirty. I have also made some models, stepped on some pitfalls, and have a more specific understanding of normal force analysis than before. So I want to write a simple summary first to record what I have done in this recent stage, some basic understandings, and the problems encountered in the analysis. On the one hand, it is convenient for you to look back later, and on the other hand, it is also a staged arrangement for your study during this period.
1. Why did we start doing normal force analysis recently?
I have been working in the connector product engineering line before, and I am more familiar with product structure, specifications, reliability, mass production introduction, and customer problems. In the past, when problems related to terminal normal force were encountered, they were mostly judged through test results, product experience, and structural understanding. For example, whether the terminal is pressed hard enough, whether the structure is too hard, and which position may be prone to high stress and deformation. In fact, there are some empirical judgments on these before. But experience is experience after all. In many cases, you know that "there may be a problem" and you actually see the problem in the simulation are two different things. Recently, I have started to do normal force analysis on products such as DDR and PCIe. For me, a big significance is that in the past, many things were based on experience, but now I can try to use analysis to see it more concretely. To put it bluntly, it means moving from "I know something is wrong based on my feeling" to "I can see why it is wrong and where the risks probably lie."
2. What is the core of the normal force analysis that I understand now?
On the surface, normal force analysis looks at a force value, but in fact, it is not as simple as just looking at "how many grams of force this terminal presses down." I now understand that it is at least looking at these things:
First, under the target stroke, is the normal force of the terminal sufficient? This is the most direct and basic.
Second, after the terminal is pressed down, is the stress too high? Because in some structures, in order to produce normal force, the elastic arms are made harder. Although the force value increases, the high stress also increases.
Third, is the terminal deformation method reasonable? The final reaction force value of some models looks okay, but the deformation mode is actually not smooth. There will be unnatural local distortion, local excessive deformation, or abnormal contact status.
Fourth, will there be any problems with rebound later? In other words, can it bounce back healthily after being pressed? Is there any risk of permanent deformation?
So now I feel more and more that normal force analysis is not just about looking at a result value, but about judging whether the entire force state of the terminal is healthy.
3. Some basic ideas when doing DDR and PCIe normal force analysis recently
In the recent analyses, although the products are not exactly the same, the overall ideas are actually similar. Generally speaking, the main body of the terminal is built, and then the rigid body or corresponding pressure object is set up, and the process of being pressed down is simulated through displacement loading. Then based on the actual structure, we define which positions are fixed and which positions are truly constrained, and finally extract the results through contact and reaction force. It sounds simple to say, but when I actually do it, what I feel most deeply now are two things:
1. Boundary conditions must be as close to reality as possible. I understand this more and more now. Because the terminal is not "free to hang in the air" in the product, it must be restricted by conditions such as plastic, fixed feet, plastic covering area, and assembly state. If the boundary conditions are set incorrectly, the final analyzed force values and stress distribution will be easily distorted. So now when I make models, I don’t just think about “just fixing it”, but pay more attention to how it is restricted in the real product.
2. Contact issues are often more troublesome than imagined. In the normal force analysis, contact is basically unavoidable. When the indenter contacts the terminal, or the terminal contacts the corresponding mating area, this process itself will bring about a lot of nonlinearity. This is also the most obvious feeling I had during my recent analysis: many times it is not that the model cannot be built, but that it may not be easy to calculate after it is built. Especially when the initial contact state is not smooth enough, the local chamfering is too sharp, the friction settings are inappropriate, and the mesh quality is average, the model will become easily stuck.
4. After this period of time, what results will I focus on?
Recently, when I do these DDR and PCIe normal force analyses, I usually focus on the following types of results.
1. Force value results. This is the most direct. It is to see how much the reaction force is under the target displacement and whether it can meet the product requirements. But now I am less and less focused on just one value. Because sometimes the force value seems sufficient, it does not mean that the design is really reasonable.
2. High stress areas. This part is also critical. The root of the terminal, the bend transition area, the vicinity of the contact area, and the location of local geometric mutations are usually the places that need to be focused on. Because if the stress in these areas is too high, it may cause problems such as yielding, residual deformation, and reduced service life. In particular, products such as DDR and PCIe have high requirements for contact stability, so they cannot just produce normal force, but also need to see whether this force is "produced healthily."
3. Deformation state. This is something I attach great importance to now. Because on the surface, the numerical values of some models are just okay, but when you look at the deformation cloud diagram, you will find that the stress path of this terminal is not natural. This feeling of "not going well" is actually very important. Many times there are problems behind the structure, and there are already signs of problems in the deformation mode in the early stages.
4. Rebound and residual risk. If the model continues to the unloading stage, then I will also pay attention to its bounce back state. Because this step can better determine whether the elastic reserve of the terminal is sufficient. Sometimes the problem doesn't seem to be big when it is pressed down, but it is more likely to expose the problem during the rebound stage. This is also something I felt deeply during my recent analysis.
5. Some typical problems encountered recently when doing this type of analysis
After doing it recently, I feel that although normal force analysis is a relatively basic type of stress analysis, it is not easy once you actually get started.
1. Convergence problems are common. This should be considered one of the most common problems I have encountered recently. Especially after the contact is established, it is easy for the model to start to be difficult to calculate. For example: the initial contact is not smooth enough, the local chamfering of the terminal is too sharp, the friction effect is relatively large, the mesh is not ideal enough, and the local contact state changes too fast during the displacement advancement process. These things will make the solution process more painful.
2. You can pass the pressure, but you may not be able to pass the rebound. I also have a clear understanding of this issue. Some models can still run smoothly during the pressing stage, but when it comes to the unloading and rebounding stage, they are prone to errors. Because in the rebound stage, behaviors such as contact relationships, stress release, and deformation recovery will be more complex, and the nonlinearity of the model will be stronger. So many times, the real difficulty is not "suppressing it", but "how to come back after suppressing it".
3. Force values and high stress are often contradictory. This is actually a very typical problem in connector terminal analysis. If you want to increase the normal force, you usually find ways to make the structure harder; but when the structure becomes hard, the stress will often increase. Therefore, many times the analysis is not pursuing the "maximum force value", but looking for a more appropriate balance: the force value is sufficient, and the structure is not too dangerous. I think this is where normal force analysis is really valuable. It doesn’t just tell you “whether it’s enough”, but helps you see “what the price is”.
6. After doing this period of time, I have some understanding of normal force analysis.
After recently doing a normal force analysis on products such as DDR and PCIe, I have several clear feelings.
1. Normal force analysis is not just about looking at a number. When you didn't really start doing it before, you might have thought: Anyway, if you look at the normal force value at the end, isn't it almost the same? But after you actually start doing it, you will find that this is not the case. What is really valuable is when you connect the force value, stress, deformation, and rebound.
2. Simulation results must be understood in conjunction with the product. Abaqus does not automatically understand products for you. After the model is run, where are the results, how to explain them, which ones are reasonable and which ones are dangerous, ultimately depends on the understanding of the product structure and actual working conditions. So now I feel more and more that the experience I accumulated in product engineering in the past is actually very useful here. Because you don’t just know how to click on the software, but you know how this terminal works in the real product.
3. The value of CAE lies in “seeing problems in advance”. This is also what I agree with most now. In the past, many things could not be made clearer until samples, tests, and exceptions were released. Now through analysis, at least some obvious risks can be exposed in advance. Although simulation cannot completely replace testing, it can indeed make design judgments earlier and more proactive.
7. Meaning to myself
For me, I recently started doing normal force analysis on DDR and PCIe, which is more than just learning one more analysis case. More importantly, I feel that I have really started to move from product engineering to CAE. In the past, my strengths were more in product understanding, specification understanding, exception handling, and customer problem response. Now if I slowly add the stress analysis aspect, then I will have a more complete perspective on the problem later. To put it simply: In the past, I relied more on experience to deal with problems; now I am beginning to try to combine experience and analysis to judge problems in advance. I think this change is quite important.
8. Things that need to be filled in later
I am definitely still continuing to accumulate this area now, and there is still a lot to make up for in the future. For example: the difference in normal force analysis under different product structures, rebound and residual deformation will be made more solid, insertion and extraction force analysis will continue to be supplemented, analysis results will be compared with actual test results, and core difficulties such as contact/friction/convergence will continue to be explored. In general, the recent normal force analysis of DDR and PCIe is a beginning for me. It’s not very deep yet, but I have definitely made progress and started to form some understandings of my own.
ending
After actually carrying out the Abaqus normal force analysis during this period, my biggest feeling is: On the surface, this is analyzing the mechanical performance of a terminal, but in fact it is training myself to understand the risks of a product earlier. Especially products such as DDR and PCIe are sensitive to contact status and structural stability, so normal force analysis is still very meaningful. I am still continuing to do it now, and I am still continuing to step into the pits. But at least this step has begun. If I continue to accumulate in the future, I hope that I will not only be able to make products, but also be able to use analytical methods more and more to look at products deeper and earlier.
Part 2: Summary of Abaqus insertion and extraction force analysis
Recently, in addition to doing some normal force analysis of DDR and PCIe, I also started to do insertion and extraction force analysis of connector terminals. After doing this kind of analysis, my feeling is very straightforward: the insertion and extraction force analysis is more complicated and more time-consuming than the normal force analysis. Because it not only looks at the reaction force after the terminal is pressed down, but also simulates the contact, sliding and force changes of the terminal and the card during the entire insertion and extraction process. This involves friction, chamfering, contact stability, mesh, and analysis step settings, so the requirements for model details are relatively high. This article will simply make a summary and record some of my recent understanding of insertion and extraction force analysis.
1. How do I understand insertion and extraction force analysis now?
Insertion and extraction force analysis, to put it bluntly, is to see how the force of the terminal and the card changes during the insertion and extraction process. It mainly does not look at a final result value, but at the entire process: when contact begins, whether the chamfer transition is smooth, whether the peak insertion force is large, what the resistance is when pulling out, whether there are abnormal fluctuations in the force curve, whether there is any jamming or excessive local stress during the contact process. Therefore, this type of analysis will be closer to the actual use state than the normal force analysis, but it is also more difficult to do.
2. Several key points in the analysis of insertion and extraction force
1. Friction coefficient: The friction coefficient has a great influence on the insertion and extraction force. Because the surface of the terminal and the plug-in card is always sliding during the plug-in and pull-out process, friction itself is an important component of the plug-in and pull-out force. If this parameter is set too large, the insertion force may be significantly high, and the model will be more likely to fail to converge; if it is set too small, although it may be easier to run, the results may deviate from reality.
2. Analysis step stability parameters: During the plugging and unplugging process, the contact state is constantly changing, so the stability setting of the analysis step is very critical. Especially when first contacting chamfering, when contact points are switched, or when local sliding suddenly becomes stronger, the model is prone to instability. Settings such as automatic stabilization and time increment will often directly affect whether the model can run smoothly.
3. Number of analysis steps: The insertion and extraction force analysis cannot be too rough. If the number of steps is too few and each step is too large, problems such as failure to keep up with contact changes, uneven force curves, and inability to capture local peaks may easily occur. Especially when the terminal is chamfered, this section is usually the most sensitive, so the number of steps should generally be divided into finer steps.
4. Plug-in card chamfering: If the chamfering is not smooth or too sharp, problems such as too stiff contact cuts, excessive insertion force peaks, and too high local stress on the terminals may easily occur in the model. Therefore, chamfering is not only a matter of assembly experience, but also critical to the analysis results themselves.
5. Mesh fineness: Mesh is sensitive to both contact and results. If the mesh is too coarse in the contact area, chamfer transition area, terminal root, and small fillets, the contact will not be smooth and the stress may be inaccurate. However, the entire model is made very detailed, and the analysis time will be too long. Generally, it is better to be more detailed in key areas and leave other areas appropriately.
3. General process of insertion and extraction force analysis
When I do this kind of analysis now, I probably follow the following idea: first sort out the geometry of the terminals and plug-in cards -> establish materials, contacts and boundary conditions -> set the displacement loading of the insertion process -> add the extraction process according to the situation -> adjust the analysis steps and stability parameters -> check the grid, focusing on refining the contact area and high stress area -> look at the force curve, stress, deformation and contact status after running the results. Overall, the process is not particularly complicated, but the real difficulty lies in the details of each step.
4. My current common errors and their causes
After doing this recently, the common questions are mainly the following:
1. No convergence after contact: The chamfer may be too sharp, the initial contact is not smooth, the friction is too large, the mesh is too poor, or the step size is too large.
2. Pressing down can pass, but continuing to insert does not work: This means that the contact path has become complicated. It may be that a certain position is stuck, the contact switch is too sudden, the chamfering or the path is not smooth enough, or the stability parameters are weak.
3. The force curve jumps a lot: it may be that the number of steps is too few, the mesh is too thick, the contact is not smooth, or the friction settings are unreasonable. So many times, the problem is not necessarily that the model is completely wrong, but that a certain part is too sensitive.
5. Several common directions for insertion and extraction force analysis and optimization
My own recent experience is that if the insertion and extraction force analysis is difficult to calculate, or the results are not ideal, you can generally look at it from the following directions:
1. First check whether the chamfering and contact paths are smooth. In fact, many problems are not parameter problems, but the contact path of the product itself is not friendly enough.
2. Check whether the friction coefficient is reasonable. If the friction is too high, the model will often be difficult to calculate and the force will be too large.
3. Adjust analysis steps and incremental control. Making the process a little more detailed will, in many cases, be more stable.
4. Optimize the grid. Especially in the contact areas, chamfer areas, and roots, mesh quality is critical.
5. Check whether the boundary conditions are too tight or too loose. Unreasonable boundary conditions will also affect the results and convergence.
6. My overall feeling
In general, the insertion and extraction force analysis is indeed a type of analysis that is more "detailed". Friction coefficient, stability parameters, number of analysis steps, chamfering, and mesh will directly affect the results and solution process. In addition, the plugging and unplugging path itself is relatively long, so the analysis time is generally relatively long. But precisely because of its complexity, its value is even more obvious. Because it can be closer to the real card insertion process, and it is easier to expose problems in the structure in advance. I am still groping in this area, but I have become more and more aware that this type of analysis is not just a software operation problem, but also a comprehensive ability to understand the product, contact understanding, and modeling details.