老兄! 你那是不是太厚了!!??? 能说说你的板料的厚度吗?? 还有上下模的间隙是多汪??作者: jeffwxm 时间: 2005-9-23 22:37
Thanks. Actually not. The simulation I am now doing is a real project according to the production process. The
clearance between the die and punch for each side is 0.254mm, and the original blank thickness is 0.211mm. So, in
reality, the thickness of the final can should be smaller or less than 0.254 mm. I have discussed this issue with some of
the guys in working in LSTC, which is the company developed LS-DYNA, the solver of Dyanform, they still can not
provide good and clear answer.作者: dmdongmin 时间: 2005-9-29 16:35
兄弟 看来我见识太少了 敬候你的分析结果 其实看过你的贴子过后我也试了试，不过不是你那样做的，是想通过在dynaform 里生成的冲模凹模的曲面，导出到ＰＲＯ／Ｅ里做成模具，但是现在还没有成功，先在
ＳimWe里问了一下，但是现在还没有得出那间隙能否生成，精确度有多高？还在试验啊 作者: jeffwxm 时间: 2005-9-30 03:46
Hi, there,
I think creating the surface with curvature of the die and punch in Dynaform is more difficult than
doing it in CAD software, e.g. ProE, but should be successful. There are lot of questions in the
solver of Dynaform ---- LS-DYNA. Maybe I am wrong. I was trying to give different scale factor
(SLSFAC), normaly it is between 0.01 to 0.1 (default value) in "Control Parameters" of Dynaform,
such as 0.2, 0.1, 0.5, and the reason that I gave the bigger value than 0.1 came up from the
LS-DYNA Theory Manual that increasing the scale factor in penalty method could reduce the
penetration possibility when the pressure is too high between contact surfaces (slave and master), while I got relatively smaller difference between results of thickness. But with 0.2, it seems that
the bigger thickness of the final can is only 0.257795 mm, which is a little bit bigger than the
clearance 0.254mm.
I am going to have a Dynaform training next month and hopefully they could answer me the
question. And then I will put it on.作者: dmdongmin 时间: 2005-10-3 20:38
我查了查Ls-DYNA的帮助手册，我想想要解决你的问题是Dyna的接触定义问题， 我看了一下资料介绍有四十多种接触，看资料显示应该是出现了渗透，让单元滑动到了接触厚度中去了，要修改分析的参数的关键字扩充接触面作者: w5551 时间: 2005-10-3 22:29
没用过作者: jeffwxm 时间: 2005-10-4 01:44 标题: Thanks a lot to dmdongmin and everyone Hi, dmdongmin, thanks a lot for your suggestion. In the beginning, I thought it should be the problem of contact since there could only be the contact surface penetration that can lead to big thickness with big minor or major strain. So, I tried to use several different contact type including "FORMING_ONE_WAY_SURFACE_TO_SURFACE", "FORMING_SURFACE_TO_SURFACE", "ONE_WAY_SURFACE_TO_SURFACE" and "SURFACE_TO_SURFACE", but they seems not working. Since I am now
working alone on this project and got nobody to discuss with, I really don't know how to do next. According to your advice
"要修改分析的参数的关键字扩充接触面", could you please explain a little more or give me some example keywords?
Thanks a lot!作者: dmdongmin 时间: 2005-10-4 09:59
我看了一下是在K文件里的*Contact中的SST和MST两个参数可以控制，但是我也没有用过，试试吧！显示说当壳单元的厚度小于1Mm时就要增加它们的接触厚度，现在正在学习研究中作者: jeffwxm 时间: 2005-10-4 20:18 标题: Thanks a lot, dmdongmin! I will try it and let you know. 作者: hoby 时间: 2005-10-5 10:26
说的没错！作者: dmdongmin 时间: 2005-10-5 18:38
只要修改DYNAFORM里的接触参数就行了！那里的改变会修改到K文件的生成，里的SST就是 ＳＣＡＶＥ ＳＵＲＦ。ＴＨＩＮＣＫＮＥＳＳ的参数，要CONTACT PARAMETERS里有的！很多，最好查一下LS_DYNA里的说明再修改，我没有全部查到每个参数的说明，还没有一个好的数值介绍给你，查到再说吧！作者: jeffwxm 时间: 2005-10-6 03:15 标题: contact parameters set-up Yes, you are right. We could set up the parameters in Dynaform directly. In Dynaform, "Tools" ---> "Define Tools", we
could select select the particular tool and specify the contact type, for my case, "FORMING_ONE_WAY_SURFACE_TO_SURFACE". Then, under "Define Contact Parameters", there are several parameters could be specified. "SCALE FACTOR SLAVE" means SFS in Dyna; "SCALE FACTOR MASTER" means SFM; "SLAVE SURF. THICKNESS" means SST; and "MASTER SURF. THICKNESS" means MST. I am still trying. There are too many variables.
You could download all the manuals from the "Tutorials" under www.feainformation.com if needed.作者: dmdongmin 时间: 2005-10-6 19:00
谢谢！that is OK!作者: tsingtsang 时间: 2005-10-8 08:51
受教 谢谢作者: diablohellfire 时间: 2005-10-15 23:59 标题: 变厚很正常 板材拉延的时候，有的地方变薄，有的地方变厚，变薄太厉害就会拉裂，变厚太厉害就会起皱。模具间隙一般为1.1t mm,t是板材厚度。dynaform里面可以定义板材和模具之间的穿透值，允许板材变厚比模具间隙大，如果没有穿透的话，接触计算就会报错。作者: jeffwxm 时间: 2005-10-19 21:42
"有的地方变薄，有的地方变厚，变薄太厉害就会拉裂，" You are right. "变厚太厉害就会起皱" I think it is only possible when the pressure pad (binder) force is not big enough and the wrinkle will happen only at the locations of the blank
between die and binder and can not happen between die and punch with good-designed production case since the
clearance between punch and die will iron the material. "模具间隙一般为1.1t mm,t是板材厚度。" This is the CONTACT
GAP set-up in Dynaform between tools and blank (e.g. contact gap between punch and blank), which can prevent initial
penetration. And normally, we can offset the tool from another tool (e.g. from punch to die) if and only if the geometry of
punch and die is the same. Unfortunately, this is not my case.
"dynaform里面可以定义板材和模具之间的穿透值", for shell element, the Max Distance of penetration is defined as
XPENE*(thickness of shell element) when SHLTHK=1. but I am not quite sure how the contact algorithm works. (如果没有穿透的话，接触计算就会报错) Could you please explain a little more?
"允许板材变厚比模具间隙大", then neither the final results of thickness can not be trusted, nor the stress and strain
because it does not follow the reality.
Thanks a lot!作者: jeffwxm 时间: 2005-10-27 21:42 标题: Progress till now The progress is as following till now:
1. The problem may come from slave and master element/node penetration. Reason:
1.1 The dimensions of the model are too small (least possibility);
1.2 The thickness of the original blank is too small;
1.3 Mass of the final product is increased a lot with mass scaling;
1.4 Maximum penetration depth allowed is too big;
2. Methods to reduce the penetration but not to solve the problem:
2.1 Make DT2MS = 0;
2.2 SSF = 1; MSF = 1; make SLSFAC = 0.1 (or bigger enough without dynamic instability);
2.3 Make XPENE smaller than default value 4.0, but can not be too small. (I am not sure).
etc.
3. Conclusion:
3.1 LS-Dyna can not really solve this kind of problem, especially when the thickness is too small, and it has limitations.
3.2 Try to use other software. I am going to try MSC.Marc (preferred) or PAM STAMP for sheet metal forming.作者: moldwizard 时间: 2005-11-6 11:13
楼主看来是个纯理论型,呵呵.
其实DYNAFORM的帮助中都说明过,但有的只是一句带过,
楼主还需要多多品一下.同时也需多到现场去看看调试过程.
这样对做分析有很大的好处.
结果出来了其实还是要看经验去分析其结果是否合理.
如果结果不合理说明你做的不对,在去调整参数.
有经验了,算一次基本就能得出定性的结果.作者: yhs527 时间: 2005-11-9 08:37
什么样的模拟结果才是最真实的？我也同意19楼的看法，要理论实际结合！呵呵作者: yanjizi 时间: 2005-11-17 09:26 标题: 很好的hypermsh视频教程，希望版主加分！ 我怎么上传不了附件啊
老兄! 你那是不是太厚了!!??? 能说说你的板料的厚度吗?? 还有上下模的间隙是多汪??