Put the output of solve in a list:

expand(thaw([solve(subsindets({sys}, suffixed(_), freeze@``) ,{x})])); 

The solve in your Question gave output as a single piecewise expression rather than as a sequence of multiple sets. Sequence-form output usually needs to be put into a set or list before further processing.

@farah adanan Please read this MaplePrimes Question very carefully, including the paper that the Question's author attached to a Reply, my Answer "It's a BVP", and all the followups to that Answrer: https://www.mapleprimes.com/questions/224706-Please-I-Need-Some-One-To-Help-Me-With#answer249172

That Question is very similar to your Question, and the paper is very similar to the paper that you attached. Indeed, it's also about nanofluids. My extensively coded Answer, which replicates 56 plots and 2 tables from the paper, could be used for your problem with very small modifications.

@Kitonum Or, easier still, 1/M

@vv The problem is not caused by using a variable whose name begins with underscore. It is caused by using a variable, of any name, that is assumed integer.

@farah adanan In the linked paper, they used 40 for infinity.

Do you have some reason to think that it should converge for Z=4?

@Annonymouse Your error comes from remove(evalb, s). In order for it to work, s must be a list or set of relations (such as equations or inequalities) such that each can be determined to be true or false. Expressions such as polynomials that aren't relations won't work. When I used it in earlier code for you, we were dealing with lists of equations. If eq is an equation, then evalb(eq) returns true iff the left side and right side of eq are identically the same. Thus, I was using it to remove vacuous equations of the form x=x.

@Muhammad Usman I just updated the Answer so that it also displays the symbolic summation in sigma notation.

@ianmccr I'm not sure if you're saying that you still don't fully understand the syntactic distinction between evaluating a procedure and executing it. If so, I think that the key distinction is that to execute a procedure, you must give it an argument sequence (which could be the NULL (or empty) sequence). This is usually done by putting a pair of parentheses containing the argument sequence next to any reference to the procedure, evaluated or not. So, if your procedure is in table T (exported from module M) at index P, it could be executed by

M:-T[P]();

@Annonymouse Could you give an example of what you mean by "setting up the options for plot3d"?

@raskr Could you post a Maple 2017 worksheet showing this? I don't have Maple 2017 to check. But certainly in both Maple 2016 and Maple 2018---both of which are open on my desktop at this very moment---and in several earlier editions, interface(displayprecision) sets the number of digits to be displayed after the decimal point (which is ridiculous, except perhaps in a financial context) and does not affect the number of significant digits displayed. So, I'd be surprised that Maple 2017 was the lone exception.

There is a size option to plot, for example,

plot(sin(x), x= -Pi..Pi, size= [1000,500]);

The numbers refer to the number of pixels.

@Adam Ledger Ah, yes, this new series of plots is much more interesting!

This is just a mathematical aside, unrelated to your font issue, which I don't know how to solve.

Are you talking about ordinary integer arithmetic here? Then x*y is necessarily divible by gcd(x,y), so floor is redundant. Furthermore, that quotient (when nonnegative) has a standard name: least common multiple, or lcm.

@MapleMathMatt A potential problem with your approach is that if a name u has multiple layers of indices, then op(0,u) only strips off the outer layer, returning a name that's still indexed. For example,

op(0, x[a][b]);