SolidWorks UK

Applying dead loads to fatigue

I was recently asked how to apply a dead load in a fatigue study (a none varying load that is applied to a structure all of the time, as opposed to a live or dynamic load that generates an alternating stress).

The diagram below shows the definition of alternating stress (or stress amplitude) and mean stress, both of which are important in influencing the lifetime of a material.

The normal inputs to a fatigue study are the stress amplitudes, but as you can see in the next picture, mean stresses also have an impact on the lifetime of a material. The higher the mean stress, the lower the predicted lifetime.

Dead loads are not alternating but do contribute to the mean stress, so it is important that we can consider them in a fatigue study. But how do we do this?

Alternating load only case

In a normal fatigue study, the method takes the max stress, S_max from a static study loadcase and we use the loading ratio (R = S_min/S_max) to define S_min directly.

For example,

• Picking ‘zero based’ means we define S_max but S_min = 0.
• Picking ‘fully reversed’ means we define S_max and S_min = - S_max.
• Picking ‘loading ratio’ means we can define any arbitrary S_min from the S_max value.

The fatigue setup simply allows us to pick the relevant static study and along with the loading ratio this provides our alternating stress values. These are then compared to the S-N curve to get the lifetime or damage plots.

Dead load case

Dead loads however are constant and therefore by definition not alternating. This means they only apply a non-zero mean stress, which must be considered for the reasons argued above, but cannot be applied in SolidWorks Simulation in the normal way.

The method we use therefore is the ‘Find cycle peaks’ method. This method takes the peaks of multiple loads to find the damage so we must build our static studies with the ‘peaks’ in mind.

Therefore if we had one alternating stress and one dead load, we would need two studies:

• Static study 1 including S_max and the dead load.
• Static study 2 including S_min and the dead load.

You can see here that we are not using one study for S_max and implying S_min from the loading ratio, rather we are doing two separate studies with S_max in one and S_min in the other. However both include the dead load.

Finally, in the fatigue study you reference both of these studies using the ‘Find cycle peaks’ method. This then takes both S_max + dead load and S_min + dead load extremes and combines them to find the worst case for fatigue.

Technorati Tags: SolidWorks Simulation,Fatigue,Dead Loads

All buckled up!

I recently did a simulation product demo for a Norwegian customer and the question of buckling came up. I realised this was an excellent example of understanding the assumptions of the solutions you are producing plus the physical implications of your modelling methods.

We started off with a simple curved panel from a training example,

 

and used shell elements to perform a linear buckling study on a quarter of the model. You can see the edge restraint on the bottom right edge, and the other two symmetry restraints with a point load at one corner, equivalent to the centre of the full panel.

 

The output of interest from a buckling study is the buckling load factor, which was obtained by RMB on the Results folder, and was reported as 27.9. This means that the load must be increased by 27.9 times to make the panel buckle.

However this was a ‘linear’ buckling study, which assumes that the deformation occurring in the component before the onset of buckling must be small. If the panel shows large deflections before buckling occurs, then its stiffness may change. Our linear buckling study solves this in one step and has no opportunity to update the stiffness matrix, therefore giving us a potentially wrong answer.

To test this we ran a nonlinear buckling study on the same component, where the final load is applied in a series of incremental steps. This allows the stiffness matrix to be updated at each increment and so the stiffness of the panel is more realistic as it deforms.

In such a study we do not have a simple buckling load factor report, rather we can plot the displacement of the corner node where the load is applied in a Response Graph. The result looked like this:

where the curve needs to be read from right to the left since the displacement (on the x-axis) is in the negative Y direction, or downwards in our model.

What this shows is that the displacement increases as the load is applied up to a ‘snap-through’ point where it begins to change direction. This snap-through point provides the value of the incremental Load Factor at the onset of buckling and is equivalent to our buckling load factor. Here you can see that it is reading about 13.2, which is a lot less than the linear case of 27.9 above!

So you can see that with a nonlinear buckling study our model would buckle if we increased our load by 13.2 times, rather than 27.9 times; a much more sobering figure.

There is one final step that is worth considering too. Our model here is perfectly symmetrical and the solution is also symmetrical due to the way it has been constructed. However the real world is not symmetrical with manufacturing variations and real loads resulting in an asymmetrical setup.

To see what effect this has on our buckling load factors we re-ran the ‘full panel’ model with a slight offset of our load as an example of an inherent asymmetry. Here you can see a split line used to create a vertex on which the load is applied, slightly to one side of the middle of the panel.

The response graph of displacement at this point shows a different result again, however the buckling load factor is even lower, coming out around 11.7.

So you can see that buckling of ‘real-world’ structures should consider potential asymmetries that allow the structure to buckle sooner than in a symmetrical case.

All-in-all buckling is an interesting failure method and for an accurate solution you should consider both the real structure, your modelling methods, and the solution type to make sure it is as accurate as possible.

Technorati Tags: SolidWorks Simulation,Buckling

Naim Audio, creating a high-end audio component with SolidWorks

Naim Audio designs and manufactures a familiar range of audio equipment, but don’t look for Naim on the shelf at the mall electronics store. Naim’s tuners, amplifiers, CD players, loudspeakers and digital-audio converters are for high-end audiophiles, the kind of customers who can tell the difference between analog and digital sound, and who can tell if the harps in Mahler’s Third Symphony sound flat. Naim Audio’s customers want superior sound, and they’re willing to pay a premium for stereo equipment that can deliver it.

 

 

 

 

 

Challenge:

Develop a new high-end disk drive stereo component that cools itself “passively” by careful placement of internal assemblies and material choice.

Solution:

Use SolidWorks Flow Simulation software to model the airflow through the component and guide modifications that promote passive cooling.

Results:

· Reduced prototyping from as many as 10 prototypes to 2

· Final design is self-cooling up to 25° C (77 ° F)

· Hard drive manufacturer validated all analysis and testing results.

All of Naim’s products receive the same care in design and engineering, from a $30,000 CD player to the $4,000 Naim Uniti all-in-one unit. In designating Naim Uniti the best product in its class for 2009, Tone Audio magazine observed: “Naim  has packed the heart and soul of their famous NAIT 5i and CD5i onto one chassis, no cheesy class-D chip amps here.”

Naim has a policy – expressed on its Website – of not creating a new product “unless it has something genuinely new to offer and integrates perfectly into our product range.” In 2008, the sales and marketing departments decided that a hard disk player met that standard.

New product, new design challenge

Hard disk players store digitized music and play it through stereo systems. A hard disk player is just another stereo component in most regards, but the hard disk player had a complication that Naim’s other components didn’t. The hard drive itself. It presented one of the biggest challenges that Naim’s 30+ person design team faced in producing its hard disk player (HDX): how to keep the component cool enough to prevent the hard drive from failing prematurely. Naim products are sold worldwide in a variety of climates, so the HDX had to resist overheating even in tropical temperatures.

“We had to be very conscious of the temperature inside the unit. Hard drives have a temperature they like to turn at, and it reduces their life expectancy if they run too hot. The bearings wear out, for example,” said Naim Audio Mechanical Design Manager Paul Neville. “The HDX was our first product with a hard drive, and it also has significantly more software than anything else we’ve designed. A complex product means more electronics, and more electronics generate more heat. At the same time, they reduce the amount of room in the unit that allows heat to circulate through. We had to do a lot of thermal analysis before we went to market to make sure the unit wouldn’t overheat.”

Guarding against overheating required even more analysis than Naim Audio usually devotes to a new product, Neville said. Also complicating the task was Naim’s desire for the internal fan to operate only at extreme temperatures, not at regular room temperatures.

“We wanted the HDX to operate at ambient room temperature without the fan kicking in, so it had to be passively cooled,” Neville said. “That meant incorporating heat sinks into the design and using positioning of the interior components to promote airflow through the product.”

SolidWorks analysis provided fast answers

Naim has designed and analyzed its products using SolidWorks CAD, Simulation, and Product Data Management software since Neville came to the company 3½  years ago. He had worked with SolidWorks and a competing product in his previous job. Based on Naim’s size and its analysis requirements, he recommended switching to SolidWorks because of its integrated analysis tools and cost effectiveness. Also, because of SolidWorks’ growing footprint in the engineering community, Neville said switching to SolidWorks gave Naim access to a bigger talent pool to hire from in the future.

“We design all of the physical elements of our products in SolidWorks, except for PCB boards. We do those in a specialized PCB application then convert the results into a solid model. For everything else – chassis, knobs, buttons, connectors, internal components, cable accessories – we use SolidWorks CAD software. To validate our designs, we use SolidWorks Flow Simulation.”

Neville’s team used SolidWorks Flow Simulation to model the heat flow through the HDX. Flow Simulation enabled the Naim team to analyze:

• radiation heat transfer between high-temperature surfaces;
• heat flow between and around components;
• transient flow analysis for simulating unsteady flow over time;
• conduction and convection heat transfer.

HDX was the first time Naim used SolidWorks Flow Simulation in new product development. Designers used a CAD model of the HDX as a starting point to create a thermal analysis model in SolidWorks Flow Simulation. The model enabled them to predict the flow patterns and maximum temperatures inside the HDX. Designers modified the CAD models in response to the thermal analysis, then ran another analysis to ensure the modification would work.

“The SolidWorks Flow Simulation software worked very well at giving us comparative solutions,” Neville said. “For example, if we changed the material or thickness of a heat sink, we would run another analysis to see what effect that had on key components. Otherwise, we would have had to use physical prototypes, and that would have taken a lot of time. We ended up doing two prototypes to look at the thermal side of the design, but using the software we tried 10 different scenarios. If not for the software, we would have needed 10 prototypes.”

Neville said SolidWorks Flow Simulation’s ability to clone a project to compare two versions side-by-side and the setup wizards for starting tasks also helped cut time off the design process.

Analysis yields self-cooling product

Using SolidWorks Flow Simulation results as a guide, Naim settled on a design that includes a 5mm thick aluminum product case that provides a built-in heat sink for the components and a path to remove heat from product. Components were mounted to the case so their heat would radiate out of the product. Designers moved the primary hard drive close to the cover to shorten the path heat had to travel to get outside. The internal air flow keeps the fan off unless the temperature rises above 77 degrees.

The HDX went to market in late 2008 to praise from the audiophile press. “We didn't expect to be won over by a computerized box of tricks, but we have been,” said TechRadar U.K.

“Naim Audio, a name synonymous with high-end, has recently launched its very own hard drive-based music server, and on paper, it's a fairly mesmerizing piece of kit,” wrote Engadget.

Neville said Naim sent the HDX to hard drive manufacturer Seagate’s test center before it went into production, to double-check its test results.

“They confirmed all of our analysis and testing,” he said. “It’s a testament to how intuitive the SolidWorks Flow Simulation software is that we were able to achieve results like that our first time using it.”

For more information about Naim Audio’s HDX visit www.naimaudio.com

How long is the flow?

I was asked yesterday if it was possible to calculate how long the flow lasts inside the domain during a flow simulation? Think of this another way and it is easy to solve; how long does it take a particle to pass through the domain when it enters the flow inlet.

To do this, we therefore start off by performing a particle study:

1. Insert a new Particle Study
2. Define an injection of particles from the inlet
3. Run the Particle Study

Then, to find the time of a particle in the domain, simply RMB on the Particle Study created and View Results.

Hit the Summary button,

and you will get a table of results for each particle injected into the inlet.

 

The thing you are interested in here is the Residence Time, which is the “life time” of the particle, or how long it exists in the calculated domain.

However there are some other interesting details here too.

The Length is the physical distance travelled by the particle whilst in the domain. You will notice that each particle has a different length and residence time, and this is simply due to the fact that each particle will take different paths depending on their starting position in the flowstream.

The picture below shows this clearly as some of the inlet particles travel all around the circumference of the external pipe, whereas others travel more directly from inlet to outlet.

This means that you should consider if you need to know the min, max, or average of the residence times for your particular application.

You will also notice that the Summary table gives you the Fate of each particle and this leads us to another interesting option in Particle Studies.

If you go to the Boundary Conditions tab in the Particle Study setup and edit the default setup, you will see three options for the boundary conditions.

This determines what happens when the particles meet the walls of the domain, and the options are:

• Absorption - The particles are absorbed by the walls. This is typical for liquid particles.

• Ideal reflection - The particles are reflected from the walls. This is typical for solid particles.

• Reflection - Specify the normal and tangential restitution coefficients which are ratios of the absolute value of the normal and tangential velocity components correspondingly after and before the collision:

So the Fate of the individual particles may be to reach the opening, be absorbed, or possibly to end their time still within the domain when the calculation finishes.

Technorati Tags: SolidWorks Flow Simulation,Tips and Tricks,Particle Study,Residence Time

Welded to the spot – and now edge…

One of the most common questions from people who have welded structures is, “can you model welds?”. Well, obviously!

In 2010 we have a new edge weld connector, similar to the spot weld connector already available.

The tool allows you define a fillet or groove weld, either single or double sided between any two sheet metal parts or shell structures like the one below.

By picking two intersecting surfaces the edge weld is generated for you at the intersecting curve.

Now all you have to do is pick an electrode material from the list of steel and aluminium types, and provide an estimated weld size.

The weld is then ready to go!

Simply solve your simulation as usual and the forces and moments are automatically passed through the new weld onto the surrounding structure.

However the really neat bit comes next. Not only is the weld connector a new joint type, it currently compares the forces and moments through the weld to the AWS 1.1 standard and tells you if the weld size you have chosen is good enough!

To get this info, RMB the Results folder and chose Define Weld Check Plot. Hit the OK tick and you get a new dialog that lists all your welds into two groups, those that pass and those that fail according to the standard.

This allows you to browse each connector and see where you need to define a larger weld.

What is more, if you chose a weld in this dialog and hit the Details button, you get even more helpful info containing the required weld and throat size plus the forces and moments experienced by the weld in the particular configuration you tested.

And as a final benefit, you can also plot the required weld size and throat size along the weld length (from an indicated start point and direction) based upon the loads generated from the simulation.

 So you can see that with a few short steps it is possible to assess your design for weld sizes. In fact you could put any arbitrary weld size in, because the software then tells you what is ultimately needed, which is of course what you want to know!

Technorati Tags: SolidWorks Simulation,Simulation,Welds,Simulating Welds

The Future of SolidWorks Simulation?

It’s been a while since I’ve posted a simulation blog as I’ve been busy looking after my new daughter. As proof, here’s a picture showing just what a fan of SolidWorks she’s already become :-).

So enough about me, let’s get back into the swing of things with a couple of insights this week.

Firstly I received this fantastic video of a flow simulation showing a timing sequence of valves in an engine manifold.

I particularly like the simplicity with which the pressures can be switched on and off using the dependency button. Very impressive.

Up next, as part of our SolidWorks 2010 launch we have been showing some of the great new motion simulation features.

One of the main reasons for using motion simulation is to know the loads that your mechanism experiences when its moving. After all, as the mechanism moves, the loads will change at each moment in time putting varying strains on your design. How do you know if you have tested the worst case?

In 2010 we have a new tool, called stress recovery (because that is what it does). Notice the three new buttons on the motion dashboard…

The first button on the left provides a dialog allowing you to choose which components you want to see stressed, and of course which period of the motion you are interested in (perhaps the entire motion just in case you can’t decide when the worst case is!).

Having defined this and run a motion study as normal to generate the loads, the middle button    then launches the structural solver to calculate the displacements and stresses on your chosen part at each time step.

And the final button on the right let’s you pick which contour you want plotted.

So, with hardly any effort, and moving from right to left through these new buttons, it is simple to generate a proper understanding of the response of your mechanism to the changing loads generated in a motion study.

 

Technorati Tags: SolidWorks Simulation,Motion Simulation

 

Are you up to the challenge?

Challenge your knowledge on 6 common design problems:

 
Good luck and enjoy the quiz.

Its not just about Simulation you know!

Very often when encountering problems with a simulation, it is all too easy to look at the simulation inputs for a solution. For example, if the model doesn’t mesh or solve you immediately think, “is my mesh too coarse?”, “are all the parts bonded correctly?”, or even “am I using the right solver?”.

However one very important variable in the success of a simulation is the CAD model itself, which becomes very sensitive if it hasn’t been built with analysis in mind. A common example of this is when someone gives you a model converted from another CAD source or from someone who has no idea that simulation will be done on their assembly.

I came across such an example last week when working on a file that had been converted from another CAD package. The model seemed to mesh ok but when I came to solve it I got an error that read:

Now the solution to this problem is usually to do with parts not being properly restrained (as it says), leaving rigid body motion a distinct possibility.

A common next step is to run the same model through a frequency analysis, because this will happily solve for the rigid body modes and help identify where the failed connections exist. However in my model the Frequency Study wouldn’t solve either giving a similar error!

Having checked and re-checked my simulation setup for the correct boundary conditions I ended up back at the geometry, which is where the problem was finally discovered. Because various CAD software uses different tolerances during CAD creation, an imported file from another source may struggle to define global bonded contacts within SolidWorks.

To check the contacts, use Tools > Interference Detection, pick the assembly and hit the Calculate button. A list of all inter-part contacts should be given that allows you to select each one and check out the contacting areas of your model.

It is useful to note that if you have the ‘Treat coincidence as interference’ ticked and you get a series of interferences, but switching this off reveals no interferences, then the global bonded contact definition should work for you. This was the case I had in my solid model, yet I was still getting a failed solution! However going through this interference list step by step eventually highlighted the problem.

About three different regions displayed a single straight line, rather than an area as the interference region between two parts. This is shown as a thick red line (below), which is much more obvious than the other shaded areas of interference. And no matter how closely you zoom in to this line, it never enlarges to the typical shaded area.

Clearly these thick lines indicate areas of the geometry where bonding is applied

along an edge. This in turn implies a zero area for the stress to be calculated on and so an infinite result would ensue, something we can’t have.

To fix this problem I chose to create a new merged part containing all the separate components in the assembly so that the other touching faces would transmit the load and no bonded contacts would need to be used.

The moral of this story is that doing simulation needs a solid basis to work from. Don’t forget to check the integrity of your CAD model, especially if it was originally built with no concern for simulation.

Technorati Tags: SolidWorks Simulation,Tips & Tricks,Solver Errors

SolidWorks Simulation 2010

I’ve just spent the past couple of weeks with our Northern European resellers going through some of the new simulation features in 2010. Feedback has been amazing and there are some really neat technologies that I’m sure we’ll be talking about soon.

However we have also spent quite a bit of effort making the day to day tasks of using the products more simple, intuitive, and productive and so I wanted to begin by highlighting one of these.

Firstly think beam modelling, think pre-processing, and think what a pain it is when you have to select specific surfaces of a beam geometry to set up bonded contacts.

In the example above the two side members of the bridge structure support the footplate, which is modelled as one complete curved surface (highlighted below).

Therefore we need to create a bonded contact set between these three entities. In SolidWorks 2009 this would have required the user to pick the top face of each side beam to bond with the surface of the footplate, meaning a repetitive ‘Select Other’ command to drill down to the correct beam face.

This might be ok for the two side members,  but what if you had to do this for each cross member too?!

Well, in 2010 the whole process has been made more intuitive and more intelligent by allowing the user to simply pick the entire beam rather than having to select the correct face of the beam.

Here we see the bonded contact set definition requiring the two side members and the surface of the footplate alone. No mention of the beam surfaces and no need to have to drill down through layers of geometry to find them. The software automatically checks for proximity of these items and applies the bond to the correct face.

Quick, clever, and productive! Now that I like.

Technorati Tags: SolidWorks Simulation,Simulation Tips & Tricks,Beam modelling

Confusing terminology

Sometimes you come across a term in simulation that doesn’t make much sense and seems to have been put there to confuse and confound. One such term that I was asked about recently is the purpose of the Jacobian check, a mesh checking tool that I know hardly anyone uses but is one of those things that can be very useful.

The whole purpose of checking your mesh is to see if it is well defined. A highly deformed element becomes inaccurate and a mesh created with whole regions of badly defined elements will just serve to lead you astray. And where do elements tend to become deformed? In areas of highly curved geometry, which is often where you are interested in knowing the results.

So what is the Jacobian check I hear you say? Well, consider a normal tetrahedral element with mid-sided nodes.

If those mid-sided nodes lie directly between the corner nodes, then the element is not deformed. However, the moment you begin to move the mid-sided nodes away from this line, to track the edge of a piece of geometry say, then the element begins to deform.

The Jacobian check is a measure of how deformed your element has become. A perfectly straight sided element has a Jacobian of 1.0 and is ideal, and the Jacobian ratio increases as the curvature of the edges increase.

The actual calculation is done based on a series of points inside the element called Gaussian points. The user can choose how many of these points to use in the Jacobian check when they create the mesh by selecting 4, 16, 29, or the node locations as the points at which to base the calculations on.

So performing a mesh quality check for the Jacobian generates a contour plot allowing you to identify areas where you may need to increase your mesh density, replacing a few badly deformed elements with a greater number of better shaped elements.

In the example below my test component has firstly been meshed with a course mesh and then a Jacobian check has been performed. The legend reports a Jacobian ratio of 1.0 as blue and shows how the main body of the straight edged geometry can be nicely modelled with straight edged elements. But the hole in the middle requires the element edges to become curved and so the Jacobian ratio increases, here to a value of 1.59.

Simply remeshing with a fine mesh improves the maximum Jacobian ratio to 1.19 showing how more elements around the circle mean they are less mishapen.

It is generally accepted that a Jacobian ratio below about 40 is acceptable in most cases so you have to balance mesh density with run-times, perhaps refining the mesh with controls in certain areas rather than across the entire part. And as I’ve said before, always have a sanity check and ask yourself if the highest Jacobian ratio is in an area where you will see the highest stresses anyway.

Technorati Tags: Tips & Tricks,SolidWorks Simulation,Jacobian Ratio