{"id":25626,"date":"2011-11-08T20:45:00","date_gmt":"2011-11-08T20:45:00","guid":{"rendered":"https:\/\/www.padtinc.com\/2011\/11\/08\/you-dont-wanna-step-to-this-breaking-down-loadsteps-and-substeps-in-ansys-mechanical\/"},"modified":"2022-04-27T10:20:47","modified_gmt":"2022-04-27T17:20:47","slug":"you-dont-wanna-step-to-this-breaking-down-loadsteps-and-substeps-in-ansys-mechanical","status":"publish","type":"post","link":"https:\/\/www.padtinc.com\/2011\/11\/08\/you-dont-wanna-step-to-this-breaking-down-loadsteps-and-substeps-in-ansys-mechanical\/","title":{"rendered":"You don\u2019t wanna step to this: Breaking down Loadsteps and Substeps in ANSYS Mechanical"},"content":{"rendered":"\n

Did you know Warren G is still alive?!?\u00a0 I obviously confused Nate Dogg with Warren G.\u00a0 How embarrassing.\u00a0 Anyways, on to the topic at hand\u2026regulators, mount up.<\/p>\n\n\n\n

\"\" <\/picture>
Nate Dogg and Warren G using ANSYS\u2026what\u2019s next?<\/figcaption><\/figure><\/div>\n\n\n\n

So what exactly is the difference between a load-step and a sub-step?  A load-step should be thought of as a set of constraints\/loads that are being solved for while the sub-steps are how you transition from one loading environment to the next. <\/p>\n\n\n\n

\n
\n
\"\" <\/picture>
Blue = Sub-step, Red = Load-step<\/figcaption><\/figure><\/div>\n<\/div>\n\n\n\n
\n
\"\" <\/picture>
Solved for loading<\/figcaption><\/figure><\/div>\n<\/div>\n<\/div>\n\n\n\n

So as shown in the graph above, at time=1 (for a static run time is meaningless, think of it more as a % of applied load where 1=100%) the applied load is 10 pounds.  Since it\u2019s the first load-step, all applied values are ramped from 0 (by default), and there are 10-equal interval sub-steps that allow the solver to get from 0 to 10.  For load-step 2, which has an end time of 2 (load-steps must have increasing end-times), only 4 sub-steps that get progressively larger were used.  We\u2019ll get into how that works, but the important take-away is that load-step 2 starts at the end of load-step1. <\/p>\n\n\n\n

Within Mechanical, you define both load-step and sub-step controls in the \u2018Analysis Settings\u2019.<\/p>\n\n\n\n

\"\" <\/picture>
Red = Load-step, Blue = Sub-step<\/figcaption><\/figure><\/div>\n\n\n\n

By default the number of steps in an environment is 1, and the solver has control over how the loads are ramped.  As soon as you increase the number of steps you will see this reflected in the \u2018Tabular Data\u2019 window for any applied load or constraint.<\/p>\n\n\n\n

\n
\n
\"\" <\/picture>
Single Load-step Defined<\/figcaption><\/figure><\/div>\n<\/div>\n\n\n\n
\n
\"\" <\/picture>
Two Load-steps Defined<\/figcaption><\/figure><\/div>\n<\/div>\n<\/div>\n\n\n\n

When you first start defining the load for a multi-load-step analysis, Mechanical assumes that the loading will remain constant for all subsequent steps.\u00a0 This is indicated by an \u2018=\u2019 sign shown in the tabular data.<\/p>\n\n\n\n

\"\" <\/picture>
Mechanical assumes a constant value for LS2<\/figcaption><\/figure><\/div>\n\n\n\n

So after you\u2019ve gone through and defined an assembly loading process, large deflection plastic analysis to determine permanent set, or whatever else your heart desires\u2026you can go back to define how the loads are ramped from one step to the next. <\/p>\n\n\n\n

Toggling the \u2018Auto Time Stepping\u2019 to \u2018On\u2019 gives you even more options.<\/p>\n\n\n\n

\"\" <\/picture>
Auto Time Step = On<\/figcaption><\/figure><\/div>\n\n\n\n

You have the ability to define the ramping controls either by number of sub-steps or time.  Both represent the same thing, just presented through a different paradigm.<\/p>\n\n\n\n

\"\" <\/picture>
Not to be confused with a pair-of-dimes\u2026(I\u2019ll be here all night, make sure you tip the wait staff)<\/figcaption><\/figure><\/div>\n\n\n\n

Before I get into the difference between sub-step and time, I\u2019ll touch on the Initial\/Minimum\/Maximum values (note the following explanation is valid for the sub-step paradigm).  The initial value is the initial ramping value used to start off the current load step.  If the solver was able to converge easily, it can increase the load increment until it hits the \u2018minimum\u2019 sub-step value.  If the sub-step was difficult to solve, or did not solve at all, the solver can reduce the increment and try again. <\/p>\n\n\n\n

Assuming the end time of the load step is 1, here are equivalent sub-step controls (sub-step vs time) listed next to what the actual load increments are:<\/p>\n\n\n\n

\"\" <\/picture>
Equivalent Settings Assuming End Time=1, Actual Load increments shown on right<\/figcaption><\/figure><\/div>\n\n\n\n

One thing to note is that if you were to change the end time from 1 to 2, the sub-step-method would still ramp by the exact same amount for each sub-step whereas the time-method would ramp it slower.  Obviously if you are running a transient analysis these controls become very important as they determine what response frequencies you\u2019ll pickup from your structure.  The documentation recommends a time-step size of (1\/20*f), where f is the highest frequency assumed to contribute to the response.  If your time-steps are too large, you\u2019ll miss the higher frequency content that could be more damaging.<\/p>\n\n\n\n

So now you\u2019re probably thinking \u201cgreat, so now what?\u201d.  First\u2026watch that tone mister.  Second, this should be one of the first places you should turn when you run into model convergence issues.  If a model is poorly constrained you can run a first load-step with additional constraints and then \u2018release\u2019 it once the model has converged.  If you have a model held in place purely by contact you can run an initial no-applied-load-step to allow the contacts to \u2018settle down\u2019 and properly engage.  If you have a model that fails due to excessive distortion or accumulated plastic strain you can increase the number of sub-steps to ramp the loads slower.  Or you could use this information to create some fairly complex loading environments.<\/p>\n\n\n\n

I\u2019ll be going over all of this information and more (activating\/deactivating loads, command snippet use, etc) in the upcoming PADT Webinar being held on November 17th.  Hope to see you there.<\/p>\n","protected":false},"excerpt":{"rendered":"

Did you know Warren G is still alive?!?\u00a0 I obviously confused Nate Dogg with Warren G.\u00a0 How embarrassing.\u00a0 Anyways, on […]<\/p>\n","protected":false},"author":31,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2323],"tags":[],"post-podcast":[],"_links":{"self":[{"href":"https:\/\/www.padtinc.com\/wp-json\/wp\/v2\/posts\/25626"}],"collection":[{"href":"https:\/\/www.padtinc.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.padtinc.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.padtinc.com\/wp-json\/wp\/v2\/users\/31"}],"replies":[{"embeddable":true,"href":"https:\/\/www.padtinc.com\/wp-json\/wp\/v2\/comments?post=25626"}],"version-history":[{"count":0,"href":"https:\/\/www.padtinc.com\/wp-json\/wp\/v2\/posts\/25626\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.padtinc.com\/wp-json\/wp\/v2\/media?parent=25626"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.padtinc.com\/wp-json\/wp\/v2\/categories?post=25626"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.padtinc.com\/wp-json\/wp\/v2\/tags?post=25626"},{"taxonomy":"post-podcast","embeddable":true,"href":"https:\/\/www.padtinc.com\/wp-json\/wp\/v2\/post-podcast?post=25626"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}